Professional Documents
Culture Documents
Submitted by:
Lawrence Lessig
Jennifer Stisa Granick
Lauren Gelman
Christopher Sprigman
Center for Internet and Society
Stanford Law School
559 Nathan Abbott Way
Stanford, CA 94305-8610
On behalf of:
SAVE THE MUSIC
Roman Ajzen, Co-CEO
5436 Harvest Run Drive
San Diego, CA 92130
CREATIVE COMMONS
Mia Garlick
General Counsel
543 Howard Street
5th Floor
San Francisco, CA 94105-3013
Comments of:
CREATIVE COMMONS AND SAVE THE MUSIC
SAVE THE MUSIC and CREATIVE COMMONS welcome this opportunity to provide
comments to the Copyright Office, and, ultimately, to the U.S. Congress, on the problem posed
1
As discussed in greater detail in Part A(II) infra, SAVE THE MUSIC and CREATIVE
COMMONS define an “Orphan Work” as any copyrighted work that is out-of-print or otherwise
not commercially exploited, and where the rightsholder is difficult, after reasonable efforts, or
impossible to find.
1
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
SAVE THE MUSIC, a group that wants to archive a mostly orphaned genre of music,
and CREATIVE COMMONS, an organization that provides tools for copyright owners to signal
what rights they reserve and what uses they approve, strongly believe the Orphan Works
problem is a serious one—one that impedes productive uses of works and merits a legislative
response. We believe that our experiences with Orphan Works allow us to offer relevant and
useful insight into the problem the current system poses and why it cannot be solved without a
change in the law. We believe that there is a workable, fair solution to this problem that may
readily be implemented without threatening either the interests of copyright owners who wish to
prevent use of their work, or the compliance of the United States with its treaty obligations.
* * *
SAVE THE MUSIC (STM) is a project of the Internet Development Fund, a California
501(c)(3) non-profit organization dedicated to the preservation of Jewish cultural music through
its digitization and placement on the Internet. Daniel and Roman Ajzen founded STM in 1998 as
a result of research conducted for a high school family history project. In just seven years, STM
has become the leading collector of Yiddish LP's in the world, archiving over 8,000 records.
Upon receipt, STM re-masters the recordings and makes them available for streaming through its
website, where it also provides historical information, lyrics, translations, sheet music, and other
resources. STM complements its archivist activities with forums for current artists and a virtual
2
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
bulletin board to announce performances and recent releases. STM also occasionally sponsors
STM does not charge for access to the content on the web site and depends entirely upon
donations for all its activities. STM has received donations of records, labor, and money from
STM repeatedly encounters problems identifying the appropriate rightsholder for many of
the works it would like to make publicly available. For example, as stated, along with the
recordings themselves, STM has considerable copyrighted non-audio holdings such as music
sheets, lyrics, books, drawings, letters, and newspapers. Practically all of these materials were
produced within the last 75 years. Many were produced by small publishers who can no longer
be found and from whom clearance cannot be obtained. STM would like to use these materials
to provide the background knowledge and history necessary to properly understand the music.
Many of these primary sources frame the issues and context of the music far better than any
explanation or description drafted by STM can and are essential to STM’s mission.
Another problem STM faces is that most of the recordings and materials it wishes to
archive were produced overseas. For example, the vast majority of the musical recordings in
STM’s holdings are foreign works that were published before 1970 and were free of U.S.
copyrights—until the Uruguay Round Agreements Act of 1994 removed these works from the
public domain and “restored” them to copyright. Many of these musical recordings were issued
by small foreign labels that have since disappeared. As a result, many of these works are Orphan
3
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
STM is a small organization with a very small budget for its archiving work. Because of
its reliance on donations, STM does not have the financial capacity to retain an attorney to
investigate the copyright history of each of its thousands of songs.2 The efforts of lay volunteers
are constantly stymied because many of the record companies that issued Yiddish-language
recordings in the past have long since gone out of business and there is no way to determine who
currently holds those copyrights. In other words, STM often faces the unenviable choice to
either spend a large sum on an attorney hoping she can find the copyright owners or to forgo
obtaining rights entirely, thereby exposing itself to crushing liability if it uses the work. The
complexity of the process has forced STM to effectively postpone the digitalization of a large
percentage of its holdings. Thus the problem of Orphan Works has hindered STM’s mission of
documenting, preserving, and spreading Jewish cultural music and other materials and denies the
To illustrate these points, we offer these specific examples of works that STM would like
to make available, but for which it has not been able to locate the rightsholder:
• Yiddishe Lider – (“Yiddish Songs”). Yiddishe Lider, a book written in Yiddish and
published in Argentina soon after the Second World War, contains first-hand accounts of
life in the Nazi concentration camps. Some of these narratives are truly masterful pieces
and are a testament to the indestructibility of the human spirit. STM would like to use
some of the narratives to illustrate the range of emotions experienced by prisoners in the
camps. However, despite STM’s best efforts, it has been unable to locate the rightsholder
2
STM is submitting these comments jointly with CREATIVE COMMONS with generous and
one-off pro bono assistance. Given the number of orphan works STM encounters and their
complex and often international copyright history, it is not feasible for STM to secure pro bono
assistance each and every time it needs to investigate the copyright history of songs it wishes to
preserve.
4
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
for this work: the publishing house is defunct and STM could not locate any records
indicating who, if anyone, now holds the rights to the book. As a result, STM has been
unable to share its contents with the world and because of similar problems with other
works from that time-period, development of the Holocaust section of STM’s website is
on hold.
• The Partisan’s Hymn. STM would like to pair the accounts of concentration camp life in
Yiddishe Lider with perhaps the most important Yiddish song of all time, Zog nit Kayn
Mol (Never Say or The Partisan’s Hymn). According to scholars, The Partisan’s Hymn
was the anthem and rallying cry of the Jewish resistance during WWII and is an
Even today The Partisan’s Hymn is played at funerals of Jewish martyrs and
retains a fundamental position in Jewish culture. Sixty years after the end of WWII, at a
time when the Jewish people are again facing serious threats from the rise of anti-
STM envisions an entire section dedicated to this song, its performers, and its
history. The proposed section will contain numerous musical versions, lyrics, music
sheets, and considerable historical background. It will be the anchor of STM’s Holocaust
content. As one would expect for such an important song, many different versions have
been recorded over the years. Dozens of these were published overseas and are thus
5
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
subject to copyright protection under the Uruguay Round Agreements Act. In particular,
STM plans to use a version sung by the Polish Army Orchestra and produced by the Lira
Record Company in Poland just after the end of World War Two. STM possesses none
of the accompanying materials for this record, such as the album cover or slip sheet, and
consequently has very limited information with which to find the copyright owner. STM
has unsuccessfully conducted Internet and library searches for additional information,
and has been unable to even find another copy. Though it is possible additional copies
exist in Poland, STM does not have the resources to investigate further. This record is
significant because it was one of the few conciliatory efforts after the war between Jews
and Poles and demonstrates the relationship between each group’s resistance efforts. But
the prospect that STM will obtain clearance for this work and many others like it is not
good, for in addition to the normal difficulties attendant to obtaining copyright clearance,
STM now has to go through the process of identifying rightsholders in countries all over
the world. The cost, both in money and time, of locating far-flung rightsholders makes
pursuing clearance prohibitively expensive. Currently, only the largest and best funded
organizations can even afford to attempt to track down the copyright owners; for a non-
• Ludwig Satz Sings His Most Famous Yiddish Theatre Classics. STM’s mission is not
only to preserve Jewish music, but also to make clear its importance to today’s culture--
thereby ensuring its future relevance. To this end, STM plans a section to illustrate
songs from Yiddish musicals in the 1940’s, 50’s, and 60’s. Many of these songs later
served as inspirations for songs in Broadway musicals and many of the writers,
6
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
composers, and singers went on to work in Broadway. By tracing the careers of those
involved in Ludwig Satz, STM hopes to show the breadth of Yiddish culture and its
The Greater Recording Company published this recording in the United States in
1973; it is thus under copyright protection. However, despite STM’s best efforts, it has
been unable to locate the rightsholder for this recording. After failing to find any trace of
the Greater Recording Company or the rightsholder, STM attempted to find some leads in
the property’s title history, but it was to no avail. STM also spoke with several
performers who were active during the company’s existence to see if they had any
information, but that too proved unsuccessful. At this point, STM was confronted with
the choice between hiring a professional or simply not using the songs. Unfortunately,
STM had no real choice but the latter. The difficulties STM has encountered in obtaining
copyright clearance for this and similar works have led it to postpone development of its
STM is participating in this proceeding to urge the Copyright Office and Congress to
amend the current system so that the organization can make these valuable and disappearing
cultural resources available without fear that it will be sued. STM has experienced the
difficulties posed by Orphan Works and strongly believes that a more user-friendly system must
be developed. STM’s encounters with Orphan Works arise because its collection was mainly
produced by small, independent recording companies. Many of them moved in and out of the
marketplace quickly and did not leave clues regarding what happened to them or their
copyrights. Nevertheless, works created by these companies do not lose their cultural value
7
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
simply because their owner cannot be located. The current system deprives creators and
archivists like STM of the opportunity to reintroduce them to the world. With the advent of the
Internet, resuscitating old songs and introducing them into popular consciousness is now possible
but the complexities of the law prevent STM from sharing some of the bright lights of Jewish
culture. STM would happily secure permission to use these works if it could find the owners. It
is very frustrating for STM to have a specific creative vision and realizable goals yet be
restrained from pursuing them by copyright rules that benefit neither the rightsholder nor the
public at large. STM has learned that the current rules for Orphan Works impede its ability to
CREATIVE COMMONS, a 501(c)(3) nonprofit based in San Francisco, offers from its
website to the general public a set of technical and legal tools, free of charge, that empower
creators to signal how they want their works used beyond the one-size-fits-all rules built into
current copyright law, and enable users to find works where the creator has signaled that certain
One tool is a suite of standardized copyright licenses that authors, artists, and publishers
can use to virtually “stamp” their online works with an alternative copyright notice. In contrast
to the traditional copyright notice – “(c) 2005. All rights reserved.” -- the CREATIVE
COMMONS copyright notice reads “(cc) 2005. Some rights reserved.” As the language implies,
the CREATIVE COMMONS form of notice is meant to draw the average Internet user’s
attention to the fact that the stamped work is available on terms somewhat less restrictive than
8
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
current copyright law would otherwise impose. The notice itself contains a hyperlink that leads
to a short summary of the legal terms associated with the stamped work: “Anyone is free to copy
my photograph,” for example, “provided they give me credit as the artist, and provided they do
This “human-readable” legal summary then links to a fuller, more traditional – and
legally binding – copyright license, drafted by CREATIVE COMMONS and lawyers from the
Silicon Valley law firms Wilson Sonsini Goodrich & Rosati and Cooley Godward LLP. The
“lawyer-readable” license spells out in detail what freedoms the user enjoys and what conditions
The CREATIVE COMMONS “Some Rights Reserved” notice also contains mark-up
language that, though invisible to the human eye, acts as a technological notice to search
technologies. This “machine-readable” version of the copyright notice, which is expressed at the
source-code level of the web page, facilitates the proper functioning of CREATIVE
<http://search.creativecommons.org>.4
This engine scours the web for pages marked with the (cc) virtual stamp, indexes them,
and then lists results by (1) file format type, (2) keyword, and (3) the legal freedoms and
restrictions the copyright holder has associated with the work. Type “Eiffel Tower” into the
3
CREATIVE COMMONS’ tools offer creators the ability to construct a license with a range of
different attributes. The main attributes are: Attribution or no Attribution, Commercial or Non-
commercial Use, Sampling or No Sampling, Derivatives or No-Derivatives and Share-Alike
(where the derived work is required to be licensed under the same terms as the original).
4
The CC search engine is available at CREATIVE COMMONS’ home page and also via a
special search box in the Firefox Internet browser. Yahoo! has also recently released a
customized search engine to enable users to easily locate CREATIVE COMMONS-licensed
works according to their specific criteria. <http://search.yahoo.com/cc>
9
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
dialog box, for example, specify “image” format, and click the “noncommercial use” button, and
the user sees a list of all the CREATIVE COMMONS-licensed photographs of the Eiffel Tower
The one-two punch of the machine-readable copyright licenses, on the one hand, and
smart copyright search engines, on the other, vastly reduces unnecessary transaction costs
between producers who want to share some uses of their content and consumers looking for
royalty-free content and who are willing to accept certain restrictions on their use.
demand far larger than even the organization anticipated. CREATIVE COMMONS first made
the CC licenses available in December of 2002. One year later, there were over 1 million web
pages carrying the “(cc) Some Rights Reserved” notice and license, according to a Yahoo! link-
back search. Today, there are over 10 million web pages under the CC licenses. What this shows
is that a large number of creators do not believe that the restrictions that current copyright law
imposes are appropriate for their particular works. Creators that choose a CC license are
signaling that something less than full-bore copyright is best for them. And this demand for
“Some Rights Reserved” licenses is not limited to the United States. CREATIVE COMMONS
has built an international network of lawyers in over 50 countries; today it offers the CC licenses
licenses include authors and producers as varied as the Massachusetts Institute of Technology,
Rice University, the Public Library of Science, Flickr.com (home to over half a million CC-
licensed photographs), musicians David Byrne and the Beastie Boys, Brazilian Minister of
Culture Gilberto Gil, filmmaker Robert Greenwald, federal appeals court judge Richard Posner,
and more.
10
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
Creative Commons is participating in this proceeding to urge the Copyright Office and
Congress to act to fix the Orphan Works problem, because its experience suggests that a market
solution is not sufficient to solve this problem The most significant lesson arising thus far from
the experience of CREATIVE COMMONS is that clear author signaling about rights – what
rights an author wishes to reserve, and what uses an author is willing to allow – lends valuable
copyrighted works. Current copyright law is just like a “no trespassing” sign on land – a user
must assume that she cannot use a work unless she first obtains the owner’s express permission.
But for many works, the cost of seeking permission for each individual use will prevent desired
uses altogether. CC licenses allow authors to provide information to the public that is the
copyright equivalent of a sign on real property allowing certain uses – a sign, for example, that
says “welcome to our store between the hours of 8 a.m. and 5 p.m.” Thus, authors who wish to
allow certain uses may effectively communicate their desires to would-be users using a CC
license.
This is the model for a solution for Orphan Works – a solution that facilitates clear
signaling that some uses are permitted or that the copyright owner no longer cares what use is
made. However, the private, author-driven approach taken by CREATIVE COMMONS cannot
be used directly to solve the problem of Orphan Works. The CREATIVE COMMONS approach
depends upon the existence of an active rightsholder – i.e., someone who is motivated to invest
in educating herself about the “Some Rights Reserved” approach, and to utilize a CC license for
her work. By their very nature, Orphan Works lack an active rightsholder. For many works that
have never been, or are no longer, commercially exploited, rightsholders are not sufficiently
11
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
motivated or no longer exist5 to educate themselves about and use a CC license. In addition,
many rightsholders of Orphan Works lack the motivation to comply with the voluntary
registration provisions of current copyright law. As a result, would-be users of Orphan Works
often face insurmountable difficulties in identifying rightsholders and seeking permission, and
Although CREATIVE COMMONS licenses cannot solve the problem of Orphan Works,
the organization believes they do point to a solution. The problem with Orphan Works is that
potential users cannot determine if a rightsholder who cannot be found is one who has
abandoned a work to any subsequent uses or one who would disapprove a potential use of the
work. And the potentially high cost of miscalculation has a chilling effect on most small
organizations, libraries, and archives. This chilling effect on creativity and innovation is wildly
disproportionate to the benefit conferred on the very few non-active owners who would
disapprove uses of their work. A set of legal rules that would oblige rightsholders to signal
whether they wish to permit uses of works (in exchange for some statutorily-established
compensation and access to certain other remedies) or whether they wish to retain their full set of
copyrights would provide some certainty to organizations like STM. In addition, and
importantly, such a tool would likely release a flood of similar information locked up by a
system that does not now produce much useful information about ownership and what owners
desire. CREATIVE COMMONS believes that its success in creating and deploying a private
signaling tool lends significant weight to its contention that the solution to the Orphan Works
problem is a statutorily-created signaling tool that will produce information about ownership,
5
Rightsholders are often impossible to find because they are corporate entities that no longer
exist, as is likely in the case of the publisher of the song Yiddishe Lider and the Greater
Recording Company that held the rights to Ludwig Satz Sings His Most Famous Yiddish Theatre
Classics, discussed above, supra Part 1.A.
12
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
allow owners to express their intentions regarding their works, and enable uses of Orphan Works
that are unused now. The registry proposal contained in Section III of these comments describes
the changes to current copyright law that CREATIVE COMMONS believes are best suited to
* * *
otherwise not commercially exploited, and where the rightsholder is hard (after reasonable
efforts) or impossible to find. Under current U.S. law, those who wish to use an Orphan
Work—to copy it, distribute it, perform or display it, or use it as the building block of a new
work—must ask the rightsholder for permission. But the current U.S. copyright system does not
keep records of copyright ownership that are complete, current, or accessible. So would-be users
can’t find the owners of Orphan Works, even when they’d be willing to pay to use them. In
many cases the works were abandoned because they failed to produce (or no longer produced)
any income. In most cases, rightsholders, once found, are willing to have their work used, often
without compensation or for a nominal royalty. But the cost of identifying rightsholders means
This is a serious problem for at least two reasons. First, most copyrighted works are
orphans. Recent historical research suggests that during the 186-year period (from 1790 to 1976)
where U.S. copyright law required would-be copyright owners to affirmatively claim copyright,
13
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
the authors of less than half of the works otherwise eligible for copyright bothered to register
their works with the Copyright Office, or to mark published copies with copyright notice, as the
law required. All of these works moved into the public domain the moment they were published.
And, importantly, of the subset of works that were copyrighted in the first place, between 85 and
90% fell out of copyright early because rightsholders failed to re-register (i.e., to “renew”) their
copyrights, as the law required up to 1976. So the data show that historically most works were
not valuable enough for owners to provide up-to-date ownership information to facilitate use.
See Christopher Sprigman, Reform(aliz)ing Copyright, 57 Stanford Law Review 485 (2004)
(attached).
Second, the scope of our current copyright law makes the problem worse. Under current
U.S. law, copyright arises whenever creative expression is “fixed” in a tangible medium. As the
result of a series of changes to the copyright laws that began in 1976, U.S. copyright no longer
requires registration, notice, or renewal—we are now living under an “unconditional” copyright
system that attaches the full term of copyright to all works, whether copyright is relevant to an
author’s ability to exploit a particular work or not. So everything from doodlings on a pad to the
latest blockbuster movie is protected by copyright at the moment it is “fixed”, without the author
being required to take any affirmative steps to protect it. All of the works that would otherwise
have fallen out of copyright under the former “conditional” system—either because of failure to
register or give notice, or because of failure to renew—remain in copyright for the (increasingly
long) full term granted under the law. All of these works are orphans, and any filmmaker,
archivist, writer, academic, musician or other person who wishes to use any such work faces the
risk of copyright infringement liability. To avoid that risk, users need to find owners and ask for
permission. But because the copyright system doesn’t keep ownership records, owners are often
14
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
hard to find or no longer exist and obtaining permission is therefore often prohibitively
expensive or impossible.
B. The Case-by-Case Approach has Failed in Canada and Japan and Should Not Be
Adopted in the United States
SAVE THE MUSIC and CREATIVE COMMONS believe that the copyright laws should
be changed to make it easier for people to use Orphan Works. As described in the Copyright
Office’s Notice of Inquiry,6 there are, broadly speaking, two possible approaches. A typical
“case-by-case” approach would allow use of a particular work only after a would-be user
provides evidence that he has met certain criteria in searching for the work’s owner, without
result, and a panel or court finds the search “reasonable”. A “categorical” approach would set up
rules that separate Orphan Works from other works, and allow use of Orphan Works without the
SAVE THE MUSIC and CREATIVE COMMONS favor the categorical approach. The
case-by-case approach has been tried in Canada and Japan. In both those countries, persons
wishing to use an Orphan Work apply to a government panel for a license. The panel
determines, for each work for which use is sought, whether the would-be user has made a
“reasonable” attempt to locate the owner; if the search is deemed reasonable and an owner
cannot be found, the panel sets a license fee. Unfortunately, because the panel engages in case-
by-case adjudication, the standards for what constitutes a “reasonable” search, and the fee set for
6
The Copyright Office did not mention in its NOI the Preservation of Orphan Works Act that
passed the Senate last year and is again before Congress this year as part of the Family and
Entertainment Act of 2005. For the sake of completeness, STM and CREATIVE COMMONS
note that, in their view, this bill falls well short of providing any effective remedy for the Orphan
Works problem because it only entitles libraries and archives to make copies of works consistent
with Section 108 of Title 17 of the United States Code and then only within the last 20 years of
the applicable term of copyright.
15
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
the license, vary. The process is time-consuming and unpredictable. As a result, it is rarely
used. The Canadian system, for example, has issued only 143 licenses since 1990.
SAVE THE MUSIC and CREATIVE COMMONS recommend that Congress amend
the copyright law to require that holders of copyrights in published works who wish to retain the
full scope of remedies that current copyright law provides register their works within a 25-year
All works—except for computer software—will enjoy a 25-year period of full copyright
protection without the need to register. Because the economic life of computer software often is
much shorter than 25 years (and is almost certain to be much shorter than the full term of
copyright), rightsholders in software will be required to register their works within five years.7
For published works that are registered within the prescribed period, rightsholders will
retain for the full term of copyright the ability to obtain injunctions and actual and statutory
Failure to register within a 25-year period following publication (or five years for
computer software) does not vitiate copyright, but moves the work into “orphan” status. Once a
work is deemed an “orphan”, it may be used without the need to ask permission, and for a
nominal fee payable under a default license applicable to all Orphan Works. No injunctions are
7
The Copyright Office and Congress might wish to identify other types of works where the term
for commercial exploitation is short, and set appropriate deadlines for registering the work.
16
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
It is important to emphasize that failure to register a work in the Orphan Works Registry
would not remove the work’s copyright protection; it would, rather, serve as a signal that the
unregistered work was an orphan, and, therefore, that the rightsholder was no longer exploiting
the work through any of the channels – customized licensing, infringement damages, injunctions
– that the copyright laws currently offer. Rightsholders who fail to register their works would be
choosing to exploit their works through a lower-cost system of one-size-fits-all default licenses
with no need to identify a rightsholder and ask permission. In this lower transaction cost
environment, many uses of Orphan Works that are impossible now will be made possible.
The default license fee will be payable to an “Orphan Fund”, where owners who did not
register, and who discover uses of their work after the fact may identify themselves and claim
any monies paid to the fund for use of their works. Penalties should be imposed for false claims
A search of the registry would be enough to constitute the “reasonable” inquiry required
to determine that a work is orphan – such an approach eliminates the need for judges to set
standards case-by-case for a reasonable search and eliminates uncertainty for users.
ownership of a copyright, whether through sale, gift, or devise, the transferee will be required to
re-register, or the work will move into orphan status if the person listed in the registry is
The law should also re-install a renewal requirement at 50 years into the copyright term.
Again, the purpose of renewal would not be to vitiate copyright, but to move works in the
registry for which renewal is not sought into orphan status. Non-renewed works would be usable
according to the same rules for default licenses that apply to unregistered works.
17
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
The databases for registration and renewal need not be administered by the U.S.
Copyright Office. The copyright registration and renewal system should be structured similarly
to the Internet domain name registration system. The government’s role will be limited to
establishing the legal rules and technical protocols according to which copyright registration and
renewal occur. Private firms will be chartered to run competing databases, and to solicit
registration and renewal information. Competition will drive down costs and complexity.
requirements would also apply retrospectively, but would be implemented only after a period
register, and thereby to retain the full scope of copyright remedies. Current rightsholders would
be given a period of five years from the effective date of legislation implementing the proposal,
or 25 years from publication, whichever is greater, to register their works. Works for which
registration is not completed during this period would be deemed Orphan Works and would be
The registry system outlined above that applies to published works would create
significant burdens if it were applied to unpublished works: it would require, for example, the
author of an unpublished email to register that work if he desired to retain the full scope of
copyright remedies. For that reason, a “notice” system should apply to unpublished Orphan
Works.
For the works of natural authors, the notice system will only apply upon the death of the
author. If a death date cannot readily be determined, the law should presume an author’s death
75 years after creation of the work in question. The notice system is limited to the works of dead
18
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
authors in recognition of the privacy interests individuals have in their own unpublished writings.
For the works of corporations, the notice system will apply 10 years after a work’s
creation.
Authors and their heirs may retain their full rights in unpublished works (outside the
notice system) by registering them any time before three years following the author’s death, or
during the 10-year period applicable to corporate works. Rightsholders in unpublished but
registered works will retain the full set of copyright remedies offered under current law.
For unpublished and unregistered works, a would-be user shall be entitled to make a use
if he (1) confirms the death of the author (or that the date of the work’s creation is within the
statutory presumption) for the works of natural authors, or the date of the work’s creation for the
works of corporate authors; (2) confirms the expiration of the three-year period for registration
for the works of natural authors or the 10-year period for the works of corporate authors; and (3)
posts a notice of intent to use for a period of six months in a centrally-administered “Claim Your
Orphan” website (such postings will include a capsule description of the work, and an image of a
portion of the work sufficient to permit recognition – e.g., for text, an image of the title page; for
film, a still of a title frame, or a short piece of streaming video). From time to time, the titles and
capsule descriptions of the works noticed on the Claim Your Orphan website shall be published
Use may commence upon the expiration of the notice period. A nominal fee shall be
payable under the same default license rules that apply to published works.
19
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
Finally, none of the changes to copyright law proposed here would create a prohibited
copyright formality under either the Berne Convention or the TRIPs Accord, and therefore
nothing suggested in this proposal would affect the compliance of the United States with these
international agreements.
Both the Berne Convention and the TRIPs Accord prohibit the imposition of formalities
on the works of foreign authors if those formalities interfere with the “enjoyment and exercise”
of copyright. Nothing contained in this proposal would interfere with copyright’s “enjoyment
and exercise”. Failure to register a work in the Orphan Works Registry would not remove the
work’s copyright protection; it would, rather, serve as a signal that the unregistered work was an
orphan, and, therefore, that the rightsholder was no longer exploiting the work through any of the
channels – customized licensing, infringement damages, injunctions – that the copyright laws
currently offer. Rightsholders who fail to register their works would be choosing to exploit their
works through a lower-cost system of one-size-fits-all default licenses with no need to identify a
rightsholder and ask permission. In this lower transaction cost environment, many uses of
Importantly, whether the default license applies to any particular work is within the
lower-cost default license, rather than the very high-cost rules that current copyright law imposes
on all rightsholders, is not a forfeiture of rights. And it does not detract from the United States’
promise under the Berne Convention and the TRIPs Accord to protect the enjoyment and
exercise of the copyrights of foreign authors. If anything, the creation of an Orphan Works
Registry would promote the enjoyment and exercise of copyright by allowing rightsholders
whose works were unable find a market under current high transaction cost copyright rules to
20
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
find a market in a lower transaction cost environment that better suits the economics of most
creative works.
For a fuller explanation of why the Orphan Works Registry, and the default licenses
associated with it, do not detract from U.S. adherence to the Berne Convention or the TRIPs
Accord, see Christopher Sprigman, Reform(aliz)ing Copyright, 57 Stanford Law Review 485,
V. CONCLUSION
The Orphan Works problem is a serious one that has a significant impact on access to
information for purposes of research, teaching, learning, and artistic creativity. SAVE THE
MUSIC and CREATIVE COMMONS believe this proposal will solve the problem without
affecting U.S. treaty obligations. It will allow SAVE THE MUSIC to make available culturally
significant materials that are of current public interest without incurring substantial costs to track
down owners and without fear of litigation. A change of law is needed because the problem lies
with inactive rightsholders who are by definition, not likely to opt in to a voluntary system. And,
as the experience of CREATIVE COMMONS has illustrated, a signaling solution where creators
opt out of copyright’s “all rights reserved” in favor of “Some Rights Reserved” has enabled
many thousands of authors to more effectively communicate the scope of permitted uses to be
made of their works. Such private arrangements cannot solve the Orphan Works problem, but
they do point the way toward a solution – a statutorily-established signaling mechanism, which
21
SAVE THE MUSIC/CREATIVE COMMONS “Orphan Works” Comments
Again, we commend the Copyright Office for conducting this important inquiry into the
problem of Orphan Works, and we look forward to the opportunity to discuss our proposal with
you further.
22
SPRIGMAN FINAL 12/17/2004 3:36 PM
REFORM(ALIZ)ING COPYRIGHT
Christopher Sprigman*
INTRODUCTION...................................................................................................... 486
I. THE TRADITIONAL CONTOURS OF “CONDITIONAL” COPYRIGHT ....................... 491
A. Formalities in the Early Copyright Statutes ................................................ 491
B. From Conditional to Unconditional Copyright ........................................... 494
1. Voluntary registration and notice............................................................. 494
2. Renewal .................................................................................................... 498
II. FORMALITIES AND THE “TRADITIONAL CONTOURS” OF CONDITIONAL COPYRIGHT
.............................................................................................................................. 500
A. Recording Ownership .................................................................................. 500
1. “Signaling” .............................................................................................. 501
2. Maximizing private incentives .................................................................. 501
B. Formalities as a Copyright “Filter”............................................................ 502
1. Registration and notice............................................................................. 502
2. Renewal .................................................................................................... 519
3. Effect of the renewal formality on the real term of copyright................... 521
4. The costs of copyright............................................................................... 523
C. Formalities and “Utilitarian” Copyright .................................................... 528
1. The Intellectual Property Clause.............................................................. 529
D. Copyright’s Increasingly Uneasy Fit with the Constitution ........................ 534
1. Term extension without renewal filter ...................................................... 536
2. Formalities as a buffer between copyright and the First Amendment ...... 537
E. Unconditional Copyright and U.S. Accession to the Berne Convention ...... 539
1. The Berne Convention .............................................................................. 539
2. Berne’s rule against formalities ............................................................... 541
3. Berne’s “practical hostility” to formalities.............................................. 543
III. REFORMALIZING COPYRIGHT .......................................................................... 545
A. Defining “Interoperable” Formalities in the Berne Convention................. 545
*
Fellow, Center for Internet and Society, Stanford Law School. The author would like
to thank Lawrence Lessig, Richard Epstein, Terry Fisher, Graeme Dinwoodie, Glenn Brown,
Lauren Gelman, Jennifer Granick, Elizabeth Rader, Mark Lemley, Laurence Helfer, Abner
Greene, Brett Frischmann, Yuko Noguchi, Goodwin Liu, Kal Raustiala, Justin Hughes, Jim
Pastore, and the participants in the Center for Internet and Society’s Speaker Series for their
helpful discussions and comments. Special thanks to Robert Harlan, Sonia Moss, Anthony
Bliss, Don Krummel, and, especially, Darien Shanske and Joe Gratz for their substantial
assistance with historical materials.
485
SPRIGMAN FINAL 12/17/2004 3:36 PM
INTRODUCTION
The recent debate between those who oppose the current trend of
expanding the duration and breadth of copyright control over creative works
and those who welcome it has focused on large and abstract questions like the
optimal duration of copyright,1 whether extension of subsisting copyrights is
constitutional,2 the degree to which technology has either facilitated or
inhibited control of copyrighted content,3 and the effect that such control has
on free speech, the public domain, and future creativity.4 The debate has
produced an insightful literature and a few creative (but thus far unsuccessful)
lawsuits.5 It has not, however, substantially altered the direction of recent
1. See, e.g., Brief of Amici Curiae George A. Akerlof et al., Eldred v. Ashcroft, 537
U.S. 186 (2003) (No. 01-618) [hereinafter Economists’ Brief]; WILLIAM M. LANDES &
RICHARD A. POSNER, THE ECONOMIC STRUCTURE OF INTELLECTUAL PROPERTY LAW 210-53
(2003).
2. See, e.g., Richard A. Epstein, The Dubious Constitutionality of the Copyright Term
Extension Act, 36 LOY. L.A. L. REV. 123 (2002); Tyler T. Ochoa, Patent and Copyright
Term Extension and the Constitution: A Historical Perspective, 49 J. COPYRIGHT SOC’Y
U.S.A. 19 (2001); Shira Perlmutter, Participation in the International Copyright System as a
Means to Promote the Progress of Science and Useful Arts, 36 LOY. L.A. L. REV. 323
(2002).
3. See, e.g., LAWRENCE LESSIG, CODE AND OTHER LAWS OF CYBERSPACE (1999); James
Boyle, The Second Enclosure Movement and the Construction of the Public Domain, 66
LAW & CONTEMP. PROBS. 33 (2003).
4. See, e.g., LAWRENCE LESSIG, FREE CULTURE (2004) [hereinafter LESSIG, FREE
CULTURE]; LAWRENCE LESSIG, THE FUTURE OF IDEAS (2001); Tom W. Bell, Indelicate
Imbalancing in Copyright and Patent Law, in COPY FIGHTS 1 (Adam Thierer & Clyde
Wayne Crews eds., 2002); Robert P. Merges, One Hundred Years of Solicitude: Intellectual
Property Law 1900-2000, 88 CAL. L. REV. 2187 (2000); R. Polk Wagner, Information Wants
to Be Free: Intellectual Property and the Mythologies of Control, 103 COLUM. L. REV. 995
(2003).
5. See Eldred v. Ashcroft, 537 U.S. 186 (2003); Complaint, Kahle v. Ashcroft, No.
C04-01127 (N.D. Cal. filed Mar. 22, 2004), available at http://cyberlaw.stanford.edu/about/
cases/Civil%20Complaint%203-22-04.pdf (last visited Oct. 4, 2004) [hereinafter Kahle
SPRIGMAN FINAL 12/17/2004 3:36 PM
Complaint]; First Amended Complaint, Golan v. Ashcroft, No. 01-B-1854 (D. Colo. filed
Feb. 18, 2003), available at http://cyberlaw.stanford.edu/about/cases/Amended%20
Complaint.pdf (last visited Oct. 4, 2004).
6. Pub. L. No. 94-553, 90 Stat. 2541 [hereinafter 1976 Act].
7. Pub. L. No. 100-568, 102 Stat. 2853 [hereinafter BCIA].
8. This Act is part of the larger Copyright Amendments Act of 1992. See Copyright
Amendments Act of 1992 §§ 101-102, Pub. L. No. 102-307, 106 Stat. 264, 264-66 (1992)
[hereinafter CAA].
9. Pub. L. No. 105-298, 112 Stat. 2827 (1998) [hereinafter CTEA].
SPRIGMAN FINAL 12/17/2004 3:36 PM
Viewed from the perspective of our current law, where copyright arises the
moment a piece of creative expression is fixed in a tangible medium,18 the
early U.S. copyright laws are remarkable for the variety of hurdles that an
author was made to clear to gain and maintain the protection of the law. The
Copyright Act of 1790,19 the first statute enacted under the authority
established in the Constitution’s Intellectual Property Clause,20 granted rights
only to U.S. authors (and their executors, administrators, and assigns) limited to
the “printing, reprinting, publishing and vending” of their maps, charts, and
books.21 The term of protection was quite short: the term of copyright in the
1790 Act was fourteen years, with a fourteen-year renewal if the author
survived to the end of the initial term.22 Most importantly for present purposes,
the 1790 Act required compliance with a fairly demanding series of formalities,
23. Id. § 3.
24. See William J. Maher, Copyright Term, Retrospective Extension, and the Copyright
Law of 1790 in Historical Context, 49 J. COPYRIGHT SOC’Y U.S.A. 1021, 1023 (2002).
Interestingly, Maher’s examination of bibliographic records from the period 1790 to 1800
reveals that only twelve works subject to preexisting state copyright law were reregistered
under the 1790 Act—a tiny fraction (1.74%) of the works registered during that period. Id. at
1025. Moreover, at the time of their reregistration, only three of the works dated from before
1784. Id. Although nine of the twelve states that had copyright laws prior to 1790 required
works to be registered as a condition of protection, most of the state registration records from
this period have been lost, so it is impossible to say how many works were registered under
state copyright and were eligible for reregistration under the 1790 Act. See G. Thomas
Tanselle, Copyright Records and the Bibliographer, 22 STUD. BIBLIOGRAPHY 77, 82-84
(1969). Nonetheless, the very low absolute number of works reregistered following the 1790
Act (i.e., reregistration of works that were previously subject to state copyrights that were
preempted by the 1790 Act) is consistent with much more expansive recent data, discussed
infra at text accompanying notes 124-134, showing that copyrighted works have, on average,
a short commercial life before their value is fully depreciated.
25. 1790 Act, supra note 19, § 3.
26. Act of Apr. 29, 1802, ch. 36, 2 Stat. 171 (1802).
27. Id. § 1.
28. Id.
29. 1790 Act, supra note 19, § 4. Responsibility for accepting registration and deposit
was later moved to the Librarian of Congress. See Act of July 8, 1870, ch. 230, §§ 85, 109-
110, 16 Stat. 198, 212, 215 (1870).
SPRIGMAN FINAL 12/17/2004 3:36 PM
And there copyright came to rest, until its major revision—and the beginning of
the move from conditional to unconditional copyright—in 1976.
42. The 1909 Act allowed protection to attach upon “publication of the work with the
notice of copyright,” id. § 12, so following 1909 it was publication with notice, rather than
registration, that served as the formality that gave rise to copyright. Following publication,
however, the Act required that the author “promptly” deposit copies of the work with the
Copyright Office, id. § 13, and, although the statute is not clear on this point, Professor
Nimmer has noted that authors were required to submit an application for registration along
with the deposit. 2 NIMMER & NIMMER, supra note 13, § 7.16[A][2][b]. Rightsholders were
barred from bringing a lawsuit for infringement of the copyright until they had complied
with the registration and deposit formalities. 1909 Act, supra note 32, § 12. In addition, the
Register of Copyrights was authorized to make a demand for deposit; failure to promptly
comply (within three months from any part of the United States except for “outlying
territorial possessions,” and within six months from anywhere else) would result in fines and
the voiding of the copyright. Id. § 13.
43. 17 U.S.C. § 410(c) (2000).
44. Id. § 411.
45. See Final Report of the Ad Hoc Working Group on U.S. Adherence to the Berne
Convention, 10 COLUM.-VLA J.L. & ARTS 513, 572-73 (1986) [hereinafter Final Report].
46. 17 U.S.C. § 412 (2000).
SPRIGMAN FINAL 12/17/2004 3:36 PM
disallows the defense of innocent infringement for works imprinted with notice
of copyright.47
The law provides a somewhat weaker scheme to incent voluntary
recordation of transfers of copyright ownership—recordation of transfers
creates a presumption of constructive notice but is not a prerequisite to an
infringement action, and failure to record does not limit infringement
damages.48
However substantial these inducements may be for owners of valuable
copyrights who foresee the possibility of infringement litigation, they are not a
replacement for mandatory formalities. The current system of voluntary
formalities creates no incentive for compliance for the large number of
rightsholders who do not expect their works to produce significant revenue. For
these rightsholders, any disadvantage that noncompliance may create in
infringement litigation is irrelevant.
Data on the rate of copyright registration confirms what logic suggests.
Figure 1 graphs the annual number of registrations for the period 1910-2000.49
This graph, which has been taken from a study of U.S. Copyright Office
data by William Landes and Richard Posner, shows that the gross number of
registrations had been increasing sharply from the end of World War I through
1991. After 1991, however, the number of registrations stabilized at a level
approximately twenty percent lower than that reached in 1991, despite very
significant growth in the overall economy between that year and 2000 (a rate of
growth that doubtless was mirrored, if not exceeded, by the increase in the
nation’s “expressive output”).50 The number of registrations should have
continued to grow after 1991, perhaps at an even greater rate than it had in the
prior decades, yet it dropped in 1992 and has failed to increase since.
Some portion of these missing registrations is comprised of authors who,
because they see no realistic prospect of commercial return from their works
and do not foresee infringement litigation, are not moved by the law’s current
inducements to register. Under the pre-1976 conditional copyright system, their
works would have moved immediately into the public domain, where they were
usable without the need to ask permission and could potentially serve as
building blocks for future works that might find commercial success. In our
post-1976 unconditional regime, however, many of these works are effectively
dead. They are copyrighted, and therefore are usable only with permission. But
the cost of obtaining permission is far from trivial. The would-be user first must
locate a rightsholder, and then negotiate for rights. The cost of negotiating a
license may be high when neither the licensor nor the licensee has any
information from other market transactions that would help establish the value
of a license. But many would-be users will never get to the negotiation stage:
the cost of identifying rightsholders, without the benefit of a registry, and often
without any reliable indication of current ownership from the work itself (either
because the work is not marked with notice or because rights have been
transferred without recordation), will often be enough to deter the use.
Perhaps the best illustration of the difficulties users face in identifying
rightsholders is the admission of the major record companies in the Napster
litigation that they were unable to produce a complete list of the copyrighted
works they claimed to own.51 Similarly, a visit to the website of the Harry Fox
Agency, the organization set up by the music publishing industry to administer
the collective licensing of musical copyrights, reveals that the agency has lost
track of hundreds of music publishing companies to whom it may owe
royalties.52 If record companies and the music publishing industry’s own
licensing agency are unable to quickly and cheaply identify the rightsholders to
whom they should be sending royalty checks, the overall cost to users of doing
so, especially in the case of works that are not commercially successful, is
likely to be substantial.
Indeed, it is not surprising that the record companies apparently invest so
little in maintaining careful records; many of the works that they own are worth
too little, in terms of expected future revenues, to merit the expense required to
keep track of them. The situation in books is probably worse: a study by Jason
Schultz of data in annual book catalogs suggests that only a tiny fraction of the
total number of books ever published is still in print—“for example, of 10,027
books published in the U.S. in 1930, only 174 [i.e., 1.7%] were still in print in
2001.”53 Publishing companies with enormous back-catalogs of out-of-print
books may find that the cost of negotiating licenses for many uses outweighs
expected revenues. Their back-catalogs are, therefore, effectively dead.
In sum, the transaction costs imposed by a system of unconditional
copyright prevent many uses that may otherwise have been made. For
unregistered works—and probably for many registered works as well54—the
current system imposes costs without producing countervailing benefits in the
form of revenues to rightsholders.
51. See A&M Records, Inc. v. Napster, Inc., 114 F. Supp. 2d 896, 925 (N.D. Cal.
2000) (recounting record companies’ contention that “it would be burdensome or even
impossible to identify all of the copyrighted music they own”), aff’d in part, rev’d in part,
239 F.3d 1004 (9th Cir. 2001).
52. See Harry Fox Agency, Inc., HFA Is Looking for These Publishers, at
http://harryfox.com/public/infoUpdateList.jsp (last visited Oct. 30, 2004).
53. LANDES & POSNER, supra note 1, at 212 (explaining results of study).
54. See infra p. 501.
SPRIGMAN FINAL 12/17/2004 3:36 PM
2. Renewal
Unlike registration and notice, which live on as ghosts, the renewal term is
well and truly dead. Under current law, all works dating from 1978 or later are
protected for a “unified” term, which is currently set for individual works at life
of the author plus seventy years, and for corporate and anonymous works at
ninety-five years.55
Figure 2 graphs the annual number of copyright renewals for the period
1910-2000; Figure 3 graphs the annual rate of copyright renewal—i.e., the
number of renewals each year as a percentage of the total number of works for
which the initial term was due to expire.56 Both graphs show what one would
expect: after renewal for pre-1978 works became automatic in 1992,57 both the
total number of renewals and the rate of renewal plummeted.58
56. These graphs are, like Figure 1, drawn from LANDES & POSNER, supra note 1, at
236, and are based on Copyright Office data collected by those authors.
57. CAA, supra note 8, § 102.
58. That renewals did not fall to zero is due to provisions of the North American Free
Trade Agreement Implementation Act, Pub. L. No. 103-182, 107 Stat. 2057 (1993), and the
Uruguay Round Agreements Act (URAA) § 514, Pub. L. No. 103-465, 108 Stat. 4809, 4976
(1994), which together restored the copyright of certain foreign works that had fallen into the
public domain for failure to comply with mandatory formalities. These foreign works are
restored “for the remainder of the term of copyright that the work would have otherwise
been granted in the United States if the work never entered the public domain in the United
States.” 17 U.S.C. § 104A(a)(1)(B) (2000).
SPRIGMAN FINAL 12/17/2004 3:36 PM
A. Recording Ownership
in each of them may, absent the workings of the law, freely be copied and
ownership of copies transferred. Therefore, although the question of who owns
a particular copy of a book presents no more difficulty than does ownership of
any particular piece of personal property, the question of who owns rights in
the expression contained in the book most often cannot be answered simply by
understanding who owns the book. The registration, notice, and recordation
formalities created the information about ownership that mere possession of a
copyrighted work could not.
1. “Signaling”
Ownership in the copyright context may be less transparent than in the case
of personal or real property, but the question of ownership is nonetheless a very
important one to our copyright system for at least two reasons. The first is that
the rights granted by the copyright laws are, unlike rights in most other forms
of property, temporary. Of the various limitations that the Constitution’s
Intellectual Property Clause imposes on Congress’s power to grant copyrights
and patents, none is more visible than the dictate that exclusive rights may be
granted only for “limited times.” This limitation, and others in the clause, are
designed to balance the need to give authors and artists incentives to create
with the equally important imperative, in a society committed to free
expression, that public access to creative works not be impeded by government-
granted monopoly. So the author (or his assign) is given a period of exclusive
ownership during which he is free to profit from his work to whatever extent
his exclusive right will allow. At the end of this period, however, the work
remits to the public, making it available as the raw material for future acts of
creativity.
Of course, if this balancing act is to work, would-be users need to know
when ownership of a copyrighted work began and when it will end. By creating
information about ownership and the term of protection—both at the inception
of copyright (registration, notice), and later (recordation, renewal)—copyright
formalities fulfilled the important function of signaling that works had moved
from the private market to the public domain. As such, formalities were
important in ensuring that the “limited times” requirement was operative at the
level of individual works.
Ownership is also important for reasons that have to do not so much with
maintaining the copyright balance, but with fully realizing the first element of
that balance—i.e., copyright’s role in spurring creation of new works. Often,
copyright owners profit by allowing others to exploit their works through
licensing arrangements. Indeed, because exploitation of creative works often
SPRIGMAN FINAL 12/17/2004 3:36 PM
Until the 1976 Act, the registration and notice requirements served as
initial conditions for which noncompliance meant copyright either did not arise
or was unenforceable. Although these initial obligations were easily satisfied,
many published works were neither registered nor marked with copyright
notice, indicating that the authors did not desire the protection that the
copyright laws would otherwise provide—i.e., that they did not project a net
present value for royalty revenue from their work that exceeded the relatively
trivial costs of complying with the formalities. Thus the registration and notice
formalities imposed an initial filter separating works with significant potential
commercial value for which authors desired protection from other works for
which protection was irrelevant. The latter class of works moved immediately
into the public domain, where it was freely usable by others (most importantly
as the building material for new works) without fee or the need to ask
permission. Furthermore, because of the notice formality, the public domain
status of many works was readily recognizable under the pre-1976 rules, even
without the need to consult a registry.
How important was this initial filter? In an age where a popular book or
record can return many millions of dollars for its copyright owner, it may be
difficult to understand why any creator would fail to comply with inexpensive
and relatively simple formalities and allow a work eligible for copyright to fall
into the public domain. But there are hints in the historical record suggesting
SPRIGMAN FINAL 12/17/2004 3:36 PM
that noncompliance was common. This next section looks at the available
historical data and attempts to quantify the effect of the initial filter.
Data on the effect of the registration formality: 1790-1870. Prior to the
1976 Act, all unpublished material was subject to perpetual common law
copyright. The relevant question, then, is what percentage of published material
was registered and made subject to copyright. The fact that a particular work
was published suggests that it had some value, and one would assume that a
relatively significant percentage of published works would be registered for the
purpose of gaining protection under federal copyright. Yet a 1987 study by
James Gilreath and Elizabeth Carter Wills of records assembled by the Library
of Congress found that of the more than 15,000 maps, charts, and books
(including pamphlets) that were published in the United States between 1790
and 1800, only 779 were registered and thus protected by copyright—a
registration rate of approximately 5%.59 William Maher, in a recent
reexamination of the Gilreath/Wills study, finds that the earlier study made
several errors that inflated the registration rate. Maher’s recalculation, using
more complete bibliographic materials, suggests that the registration rate during
that early period was even lower—3.28%.60
The Gilreath/Wills and Maher studies may both understate the rate of
registration. Both studies count some published works from this period that
were of foreign origin, and thus not eligible for protection under the 1790 Act.
And neither study accounts for the loss of some early copyright records.61 But
even if the Gilreath/Wills and Maher studies offer only rough approximations,
they suggest that a small percentage—probably at most only between 10% and
20% of works eligible for copyright protection—were registered in the decade
following the 1790 Act.62
59. See James Gilreath, Editor’s Preface to FEDERAL COPYRIGHT RECORDS 1790-1800,
at ix, ix (James Gilreath ed. & Elizabeth Carter Wills compiler, 1987).
60. See Maher, supra note 24, at 1024 n.8. Using additional information obtained from
ROGER BRISTOL, SUPPLEMENT TO CHARLES EVANS’ AMERICAN BIBLIOGRAPHY (1970), about
the number of works published in the United States, and correcting for duplicate copyright
registrations in Gilreath’s records, Maher arrives at a larger number of published works
(20,829) and a smaller number of copyright registrations (684).
61. Gilreath, supra note 59, at ix; Maher, supra note 24, at 1023-27. Further, in relying
on the number of total texts found in CHARLES EVANS, AMERICAN BIBLIOGRAPHY (1941),
Maher is repeating some instances where Evans double-counted and included foreign texts.
The recent online version of American Bibliography (available at
http://infoweb.newsbank.com to users with a password), which includes the supplement to
American Bibliography (a password-protected description of which is available at
http://infoweb.newsbank.com/iw-search/we/Evans/?p_action=help), provides a lower total
number for texts produced in 1790-1800 (12,303) than Maher’s.
62. There is some anecdotal evidence that the percentage of published works subject to
copyright remained low, although probably not as low as at copyright’s inception. In a 1961
report to Congress, the Register of Copyrights stated that “most of the great mass of
published material” did not bear a copyright notice, indicating that many authors were
SPRIGMAN FINAL 12/17/2004 3:36 PM
I could not find any existing study that calculates a rate of copyright
registration for any period following 1800. I therefore attempted my own post-
1800 calculation. What follows is a study of data regarding the registration rate
through 1870, which is the latest date for which data exists allowing estimates
of both total copyright registrations (the numerator in my calculation) and total
published copyrightable works (the denominator).63
Construction of both the numerator and the denominator posed significant
difficulties. First, for the period between 1790 and 1870, only the copyright
registrations for 1790 through 1800 have been indexed and published.64
Though catalogued, the registrations for 1800 through 1870 have not been
precisely quantified, subdivided by type or date, or published.65 Compounding
the challenge, over the period in question (1801-1870), imprints (1803), music
(1831), and photographs (1865) were added to the types of items that could be
copyrighted.
What I did have was a well-regarded estimate of 150,000 for total
registrations for the period between 1790 and 1870.66 This number may be too
low, because some records have been lost, or too high, because not all
copyrighted items were actually published.67 Given an inability to correct for
these factors, and given that they are likely to be relatively small and at least
partially offsetting, we will use 149,221 as our numerator—150,000 total
registrations for 1790 through 1870, minus 779 registrations for 1790 through
1800—in our attempt to quantify the impact of the registration requirement for
the period from 1801 through 1870.
simply not interested in securing copyright at all. HOUSE COMM. ON THE JUDICIARY, 87TH
CONG., REPORT OF THE REGISTER OF COPYRIGHTS ON THE GENERAL REVISION OF THE U.S.
COPYRIGHT LAW 62 (Comm. Print 1961); see also H.R. REP. NO. 94-1476, at 143 (1976)
(explaining that the copyright notice requirement serves four principal functions: “(1) It has
the effect of placing in the public domain a substantial body of published material that no
one is interested in copyrighting; (2) It informs the public as to whether a particular work is
copyrighted; (3) It identifies the copyright owner; and (4) It shows the date of publication.”).
63. I received invaluable assistance from Joe Gratz and Darien Shanske in conducting
historical research for this Article. Though the discussion of our findings makes reference to
the author in the singular, this stylistic convention is adopted to ensure clarity and ease of
reading, and not to suggest that the author conducted the research independently.
64. This was done in FEDERAL COPYRIGHT RECORDS 1790-1800, supra note 59.
65. See ALICE D. SCHREYER, THE HISTORY OF BOOKS: A GUIDE TO SELECTED
RESOURCES IN THE LIBRARY OF CONGRESS 94 (1987).
66. The estimate is from Martin A. Roberts, Records in the Copyright Office of the
Library of Congress Deposited by the United States District Courts, 1790-1870, 31 PROC.
BIBLIOGRAPHICAL SOC’Y AM. 81, 94 (1937). It is cited as authoritative by SCHREYER, supra
note 65, at 90; see also U.S. COPYRIGHT OFFICE, 106TH ANNUAL REPORT OF THE REGISTER OF
COPYRIGHTS FOR THE FISCAL YEAR ENDING SEPTEMBER 30, 2003, at 62 n.1 (2003); Tanselle,
supra note 24.
67. Roberts, supra note 66, at 87, 92.
SPRIGMAN FINAL 12/17/2004 3:36 PM
how much both population and GDP grew over this period—the latter made
especially significant by the extent to which the data recorded by Evans and his
successors tracks growth in both GDP and population. If instead of our
assumption of no growth in annual text production, we correlate the annual
production of texts with growth in population, then total texts leaps to 468,956,
and the percentage copyrighted falls further to 23.59%; if tied to GDP growth,
total texts would be 514,350, and the percentage copyrighted falls still further
to 22.01%.
30
15
208,986 texts in 1846
200,000
10
100,000
5
90
94
98
02
06
10
14
18
22
26
30
34
38
42
46
50
54
58
62
66
70
17
17
17
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
Year
72. For information on texts, see A CHECKLIST OF AMERICAN IMPRINTS, supra note 68;
AMERICAN BIBLIOGRAPHY: A PRELIMINARY CHECKLIST, supra note 68; EVANS, supra note
61. For population and GDP information, see Louis D. Johnston & Samuel H. Williamson,
Source Note for US GDP, 1789-2002 (2003), at http://www.eh.net/hmit/gdp/GDPsource.htm
(last visited Sept. 9, 2004).
SPRIGMAN FINAL 12/17/2004 3:36 PM
96.93 billion
$90
526,653 texts
500,000
Data for texts is an $80
estimate past 1846.
$70
400,000
395,778 texts
$60
Total Texts Produced If Texts Increase
With GDP
300,000 $50
Total Texts - Flat Lined After 1846
39.92 billion
$20
100,000
$10
$0
90 94 98 0 2 06 10 14 18 22 2 6 30 34 38 42 4 6 50 54 58 62 66 7 0
17 17 17 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18
Year
73. The information regarding U.S. GDP in Figure 5 comes from Johnston &
Williamson, supra note 72.
74. Clarence Brigham, American Booksellers’ Catalogues, 1734-1800, in ESSAYS
HONORING LAWRENCE C. WROTH 31 (1951).
75. See Mifflin v. R.H. White Co., 190 U.S. 260 (1903); Holmes v. Hurst, 174 U.S. 82
(1899).
SPRIGMAN FINAL 12/17/2004 3:36 PM
titles of periodicals, let alone the articles within them.76 There were
approximately 1200 American newspapers in 1833 and 3000 in 1860.77 As for
periodicals, there were a few hundred in 1833 and more than a thousand in
1860.78 Since there were no periodicals counted in the bibliographic collection
for 1820 through 1846, these copyrightable items are not taken into account in
the measure of the number of total imprints (i.e., in the denominator of our
calculation).79 To give some perspective of the undercount, if there were as few
as a thousand periodicals published per year on average between 1820 and
1870, and if each of these periodicals contained two copyrightable articles per
year, then periodicals alone would have accounted for all of the 150,000
copyright registrations between 1790 and 1870.
We might expect another potentially large increase in the denominator (i.e.,
another large drop in the percentage of works copyrighted) if we were able to
reliably count copyrightable commercial work product (such as labels)
produced during the relevant period. These works were copyrightable from
1803 and clearly were registered and deposited in significant enough numbers
to arouse indignation in the Librarian of Congress.80 In 1873 there were 2520
prints deposited, which shows a general awareness that these objects could be
copyrighted, but this is likely a small portion of the total output of these works
in that year, considering the country’s population was then already 43 million,
with an economy of $119 billion.
There are also factors that would tend to decrease the denominator and thus
increase our estimate of the percentage of works that were copyrighted. Evans
and his successors included government publications in their lists, which were
not copyrightable. More significantly, works by foreign authors were not
copyrightable until after 1890, and these works comprised a large portion of
American production. Here, too, there is no hard data, but there is a clear
overall trend, namely toward a majority of works being written by American
76. Evans lists newspapers and periodicals once for each year of publication, while
Shaw and Shoemaker (1801-1819) list them only once and Shoemaker et al. (1820-1846) do
not list periodicals at all. See A CHECKLIST OF AMERICAN IMPRINTS, supra note 68, at v;
AMERICAN BIBLIOGRAPHY: A PRELIMINARY CHECKLIST, supra note 68, at ix.
77. FRANK LUTHER MOTT, AMERICAN JOURNALISM 216 (rev. ed. 1950).
78. Id. Another source puts the total number of American periodicals in 1872 (just after
the end of our period) at 8110. M.B. Iwinski, La Statistique Internationale Des Imprimés,
BULLETIN DE L’INSTITUT INTERNATIONAL DE BIBLIOGRAPHIE 1, 66 (1911).
79. See A CHECKLIST OF AMERICAN IMPRINTS, supra note 68, at v.
80. Specifically, the Librarian complained in 1872 about the deposit of “printed labels,
with or without pictorial embellishment, designed for use on cigar-boxes, patent medicines,
and other articles of manufacture.” ANNUAL REPORT OF THE LIBRARIAN OF CONGRESS
EXHIBITING PROGRESS OF THE LIBRARY DURING YEAR ENDED DECEMBER 1, 1872, at 4 (1872).
This is confirmed by Roberts, who attests to seeing records of “patent medicine labels, cigar-
box linings, photographs, and all the other miscellanea which the copyright laws were called
to protect.” Roberts, supra note 66, at 92; see also Tanselle, supra note 24, at 86.
SPRIGMAN FINAL 12/17/2004 3:36 PM
authors by the 1840s.81 An assumption that 50% of total texts were of foreign
origin is conservative, as that figure is not only at the upper end of
contemporary estimates, but also does not take into account that a
disproportionately large percentage of the total number of texts in the
denominator were produced in the final two decades of the period in question—
i.e., at a time when works by American authors comprised the majority of
works published in the United States. A brief inspection of the three
bibliographic collections suggests—although an exact identification of all
foreign works is impossible from the data provided—that foreign authors do
not make up nearly half of the works recorded.82 I have not found any similar
estimates of the percentage of music publication that was the work of
foreigners. Even discounting the total number of titles (both texts and music)
by 50% and using the most conservative method for calculating the number of
imprints between 1846 and 1870 (i.e., the flatline method), 54.56%, or just
81. One contemporary source put the percentage of foreign works and reprints at 45%
between 1830 and 1842 and 30% in 1853. NIKELUS TRÜBNER, TRÜBNER’S BIBLIOGRAPHIC
GUIDE TO AMERICAN LITERATURE, at xxiii-iv (1855). Trübner’s numbers for 1853 are cited
without comment by Tebbel. 2 TEBBEL, supra note 71, at 23. Another contemporary source
puts the percentage of foreign works in 1835 at about 40%; however, this source also states
that in 1833 there were “one-third more foreign than original” books. ROSALIND REMER,
PRINTERS AND MEN OF CAPITAL 149 (1996) (citing 2 BOOKSELLER’S ADVERTISER AND
MONTHLY REGISTER OF NEW PUBLICATIONS AMERICAN AND FOREIGN 2 (1836)). Another
contemporary source for the 1850s also apparently confirmed Trübner, finding percentages
of original American titles to be roughly 61%, 58%, and 73% for the years 1853, 1854, and
1855, respectively. See CLARENCE GOHDES, AMERICAN LITERATURE IN NINETEENTH-
CENTURY ENGLAND 42 n.66 (1944) (citing SAMPSON LOW, THE AMERICAN CATALOGUE OF
BOOKS vi (1856)). Certainly the trend was toward American authors; Tebbel summarizes as
follows:
The first half of the nineteenth century witnessed the steady swing away from British to
American authors, notwithstanding the immense popularity of Scott and Dickens. In 1820,
the ratio had been thirty American to seventy British; in 1856, it was eighty American to
twenty British. While these figures may not be quite accurate, they disclose the trend
unmistakably.
1 TEBBEL, supra note 71, at 221. Tebbel follows another contemporary source here,
Goodrich, who states that the balance shifted toward American authors in the 1840s. 2 S.G.
GOODRICH, RECOLLECTIONS OF A LIFETIME 389 (1856); see also Aubert J. Clark, The
Movement for Intellectual Copyright in Nineteenth Century America 38-39 (1960)
(unpublished Ph.D. dissertation, Catholic University of America) (on file with author)
(doubting that the percentage of American bestsellers that were pirated between 1800 and
1860 was as high as 50%). For the years 1890 through 1916, books by foreigners averaged
44% (for the twenty years that complete information was collected). 2 TEBBEL, supra note
71, at 710 (citing FRED E. WOODWARD, A GRAPHIC SURVEY OF BOOK PUBLICATION, 1890-
1916 (U.S. Bureau of Educ., Bulletin No. 14, 1917)).
82. This was also confirmed by my inspection of all the California imprints calalogued
in Robert Harlan, Printing for the Instant City: San Francisco at Mid-Century, in GETTING
THE BOOKS OUT 137 (Michael Hackenberg ed., 1985). Only around one percent were of
foreign origin. No doubt this number is lower than the number nationwide, as San Francisco
publishers were clearly focused on providing works of local interest and many reprints
appeared within periodicals.
SPRIGMAN FINAL 12/17/2004 3:36 PM
more than one in two copyrightable works, were actually copyrighted for the
period 1801 through 1870.
Examination of specialized collections. This global estimate is
corroborated by evidence drawn from specialized bibliographic collections
from the end of our period. Almost two decades ago, Professor Robert Harlan
assembled and analyzed a uniquely comprehensive collection of 2571 works
published in San Francisco from 1850 through 1870.83 He found that the San
Francisco publishing industry, in a period when the book publishing industry
was much more advanced than it was for much of our broader 1801-1870 time
frame, produced 632 books (defined as works over forty-nine pages), 202
broadsheets, and 1737 pamphlets and other ephemera.84 Harlan has notes for
2053 of these works, recording whether individual works bore copyright notice.
Harlan found that only 13.48% of these published and copyrightable texts were
copyright noticed.85 The percentage is higher for certain categories, such as
collections of laws and cases (86.05%), almanacs and directories (74.42%), and
books generally (56.34%). Of all remaining items, i.e., texts of forty-nine pages
or less, only 3.95% were copyright noticed.
A review of some of the works studied by Harlan found that the works not
copyright noticed were neither all ephemeral nor obviously different from or of
less value (either commercial or cultural) than the works that were noticed. For
example, A Practical Illustration of the Movements of Hurricanes, with Plain
Directions How They May Be Avoided (eight pages, with diagrams, 1862) was
not copyright noticed, while The Law of Storms: The Various Phenomena by
Which Their Approach Can Be Ascertained with Certainty, and Practical
Directions to Mariners for the Avoidance of Their Fury (nineteen pages, 1869)
was noticed. The Miners’ Own Book: Containing Correct Illustrations and
Descriptions of the Various Modes of California Mining, Including All the
Improvements Introduced from the Earliest Day to the Present Time (thirty-two
83. Statistics provided by Professor Robert Harlan (July 28, 2004) (results of study
examining San Francisco imprints from 1850 through 1870) (on file with author).
84. Harlan, supra note 82, at 154. Harlan adds that he doubts as many as one in four
examples of job work (e.g., labels) published in San Francisco during the period has
survived, even though the number of ephemera in his collection outnumbers books by over
three to one. Id. at 145. The copyright registration records for California for the period 1851
through 1862 are available; these records indicate that only about 56% of all the works
copyrighted in California for this period (293 items) were texts. CALIFORNIA IMPRINTS, 1833-
1862, A BIBLIOGRAPHY 480-504 (Robert Greenwood ed., Seiko June Suzuki & Marjorie
Pulliam compilers, 1961). Examples of job work copyrighted include labels for the “Eureka
Compound for Fever & Ague” (1852) and “Fish’s Infallible Hair Restorative” (1861), and a
blank of a membership certificate from the Committee of Vigilance for San Francisco
(1856). So a substantial amount of copyrightable work most likely has been lost and cannot
be counted in the denominator of any registration rate calculation.
85. The percentage reflects 198 copyrighted items out of 1469 total copyrightable
items. The figure for total copyrightable material is probably too low, as it was reached using
very strict criteria to determine which items would have been copyrightable.
SPRIGMAN FINAL 12/17/2004 3:36 PM
pages, 1858) and Sketches of the Washoe Silver Mines (twenty-four pages,
1860) were not copyright noticed,86 but Guide to the Colorado Mines (sixteen
pages, including map, 1863) was noticed.
Most sermons were not copyright noticed, but liberal cleric Laurentine
Hamilton, founder of the Hamilton Free Church,87 did notice his sermons.88
Joseph Josselyn, M.D., did not notice his medical work “Man, Know Thyself!”:
A Treatise upon Sexual and Other Diseases (forty-seven pages, 1866), while
the California Illustrated Family Medical Almanac was noticed and registered
(1858). Not all works on law are copyright noticed. For example, James
William Shaw’s pamphlet Land Titles in San Francisco (sixteen pages, 1862)
was not noticed. Useful educational tools, though also often copyrighted, were
not always. For instance, Bernhard Marks’s textbook Normal Tract on
Numeration and Notation (fifteen pages, 1869), was noticed; John Martin
Spalding’s more polemical Common Schools in the United States Compared
with Those in Europe: Being a Review of the Work of Joseph Kay, Esq., on the
Results of Primary Schools in Different European Countries (thirty-two pages,
1860) was not noticed.89
Interestingly, in examining Harlan’s data, I found several works that had
been copyright noticed between 1851 and 1862 but were not recorded in the
transcribed copyright records discussed above.90 Harlan found the same
phenomenon occurring when he examined the output of one of the major
California publishers of this period, the Anton Roman firm. Of the sixty-five
works Roman published from 1860 through 1870, forty-five, or 69% of the
total, were copyright noticed, but of these Harlan could find only twenty-six, or
86. Interestingly, DeGroot’s map of the Washoe Mines was registered (item #203,
1860).
87. For background information on Laurentine Hamilton, see First Unitarian Church of
Oakland, Welcome to the First Unitarian Church of Oakland, at
http://uuoakland.org/history.htm (last visited Oct. 31, 2004).
88. Charles Wadsworth, for instance, did not notice his many sermons when published
as pamphlets, but his publisher, Anton Roman, did notice a collection of his sermons even
though at least two (and probably most, if not all) of the sermons in the collection had
already been published as pamphlets and were thus in the public domain.
89. A 1963 Practicing Law Institute monograph on copyright law by Barbara Ringer
and Paul Gitlin was published without notice and immediately entered the public domain.
See BARBARA A. RINGER & PAUL GITLIN, COPYRIGHTS (1963). Ringer was perhaps the
person most familiar with copyright formalities at that time; when the book was published,
she was Assistant Register for Examining at the U.S. Copyright Office. (She would later
serve as Register of Copyrights between 1973 and 1980.) It is exceedingly unlikely that the
chief examiner for the Copyright Office responsible for enforcing the notice formality
mistakenly omitted notice from her own monograph on copyright law, and it is likely that
many other such omissions were deliberate.
90. For example, the Marysville Directory for the Year 1855 is copyright noticed, but
not registered, even though the same publisher’s Directory for Sacramento for 1853-54 is
registered. Also, LEWIS SHEARER, A DIGEST OF THE DECISIONS AND OPINIONS OF THE
SUPREME COURT OF THE STATE OF CALIFORNIA (1859) is noticed, but not registered.
SPRIGMAN FINAL 12/17/2004 3:36 PM
40% of the total, actually registered in the records now at the Library of
Congress.91 So Roman apparently was making decisions not only regarding
which works to copyright notice, but regarding which should be actually
registered.
For the period after the centralization of copyright records in the Library of
Congress, there is definitive information on the total number of copyright
registrations by type. Unfortunately, because of the lack of bibliographic
studies, calculating total production of copyrightable items after 1846 becomes
impossible. Yet strong indications remain of the continued impact of the
registration requirement after 1846, even in an increasingly commercial
society. I conducted a survey of the holdings of the Bancroft Library at the
University of California, Berkeley, for San Francisco publications for the year
1908 and found that 20.95% of works were copyrighted, with a significantly
higher percentage of books copyrighted (34.09%) and a much smaller
percentage of the remaining, more ephemeral items copyrighted (11.48%).92
Though the San Francisco publishing industry declined considerably after
1870,93 the works that were published, including those not copyrighted, could
be of considerable political—especially local political—importance. For
instance, The Treatment of the Exempt Classes of Chinese in the United States:
A Statement from the Chinese in America by Poon Chew Ng (fifteen pages),
Shame of the Relief; Being an Expose of the Disgraceful Methods of the Relief
Committee During the Dark Days Following San Francisco’s Great Disaster
by Mary Kelly (fifteen pages), and In the Shadow of the Gallows, an Innocent
Man Condemned to Die: Resume of the Case of William Buckley, Who Will Die
91. Statistics provided by Professor Robert Harlan (July 28, 2004) (on file with
author). There are some more suggestive examples from Harlan’s data of individuals making
decisions regarding which works to notice. The California Immigrant Union copyright
noticed its informational work entitled All About California, and the Inducements to Settle
There (seventy-four pages, 1870), but not its political works, such as Arguments in Favor of
Immigration with an Explanation of the Measures Recommended by the Immigrant Union
(twenty-six pages, 1870) or Common Sense Applied to the Immigrant Question: Showing
Why the “California Immigrant Union” was Founded and What It Expects to Do (sixty-four
pages, 1869). Jacob Leon Stone does not notice his Reply to Bishop Colenso’s Attack upon
the Pentateuch (one hundred eleven pages, 1863), but does notice Slavery and the Bible; or
Slavery as Seen in its Punishment (forty-eight pages, 1863). In general, political statements
were not noticed, which makes sense, since often the whole point was to publicize a point of
view, often unpopular, such as George C. Bates’s published speech entitled, Address of Geo.
C. Bates, Esq., Which He Was Prevented from Delivering at Sacramento, on Saturday, April
19th, 1856, by a Mob.
92. For this survey I compared the records of the Bancroft Library with the published
copyright records for 1908.
93. Harlan, supra note 82, at 162. This sample is much thinner, with a total of 105
copyrightable items.
SPRIGMAN FINAL 12/17/2004 3:36 PM
94. The campaign was apparently successful, based on the later publication of another
pamphlet, The William Buckley Case: Convicted of Murder and Sentenced to Be Hanged,
Released from Prison by Governor Jas. M. Gillett on October 16, 1909 (1910). The Buckley
case was not one of a simple murder but was connected with larger labor issues, as Buckley
was accused of killing a strikebreaker.
95. Specifically, 4162 did not have a date, which is a conservative proxy for lack of
notice (the presence of a date is not tantamount to notice). Based on a sample of fifty posters,
we found no examples of a poster properly noticed without a date, though technically, post-
1909, this would have been possible for some posters.
96. LANDES & POSNER, supra note 1, at 235. Landes and Posner note that another
factor potentially contributing to the decline in registrations may be that “since 1989
registration has no longer been a condition for bringing an infringement suit for foreign
works protected [under] the Berne Convention.” Id. at 235 n.40. However, because foreign
works constitute a small percentage of works registered in the United States, the total
exemption of foreign works from voluntary registration is a relatively unimportant
determinant of registration rates.
97. Note that the consequences of failing to register a copyright are more limited post-
SPRIGMAN FINAL 12/17/2004 3:36 PM
data confirms that the number of registrations has remained flat at a level below
its 1991 peak. In 2003 the Copyright Office registered 534,122 works;98 in
2002, 521,041 works;99 in 2001, 601,659 works.100 These figures confirm that
the basic dynamic of the 1992-2000 period has persisted: unlike prior to 1991,
when registrations had been rising at a significant rate, registration growth has
ceased following the increase in fees.
The registration requirement thus encouraged authors to assess the value of
their works prior to first publication. If the author expected the work to have a
commercial value in excess of the time-adjusted cost of complying with
registration and other formalities, he would take the steps necessary to obtain
copyright protection. But if the costs of protection exceeded the expected
revenues from copyrighting, the author would not register the work.101
In sum, this initial filter separating commercially valuable works from
commercially valueless works helped focus the pre-1976 copyright regime in a
way that maximized the incentive value of copyright while reducing the social
costs. It makes sense to exclude from copyright the many works for which (in
the author’s judgment) protection is unlikely to provide more than trivial
returns, for in those cases the primary effect of copyright is to burden
subsequent use without countervailing private or public benefits.
1976. Failure to register in the post-1976 unconditional regime does not move a work into
the public domain, which raises the possibility that the decline in registrations after 1991 is
the result of many rightsholders delaying registration until an infringement occurs. It is only
after registering a work that a U.S. rightsholder may initiate an infringement suit. 17 U.S.C.
§ 411 (2000). However, failure to register at the work’s inception still imposes a penalty: a
rightsholder may recover neither statutory damages nor attorney’s fees for the period of
nonregistration. Id. § 412. Given the difficulty of proving actual damages in an infringement
suit, and the possibility that infringement may go undetected for long periods, the
rightsholder who waits until detecting infringement to register his work risks forfeiting a
substantial share of the infringement damages he might have recovered had he registered at
the beginning of the term. Accordingly, any rightsholder who, at the inception of a copyright
term, forecasts a substantial net present value for his work is likely to register. Accordingly,
the correlation between the post-1991 decline in registrations and the concomitant fee
increases holds, even though registration no longer is equivalent to complete forfeiture of
rights.
98. U.S. COPYRIGHT OFFICE, supra note 66, at 62.
99. U.S. COPYRIGHT OFFICE, 105TH ANNUAL REPORT OF THE REGISTER OF COPYRIGHTS
FOR THE FISCAL YEAR ENDING SEPTEMBER 30, 2002, at 8 (2002).
100. U.S. COPYRIGHT OFFICE, 104TH ANNUAL REPORT OF THE REGISTER OF COPYRIGHTS
FOR THE FISCAL YEAR ENDING SEPTEMBER 30, 2001, at 2 (2001).
101. If an author (or, perhaps more relevantly, a publisher) foresees little commercial
value for a particular work, that work may be neither registered nor published, in which case
the federal copyright term would not commence under pre-1976 law. See 1909 Act, supra
note 32, §§ 10-11. Rather, the work would have been subject to perpetual common law
copyright. However, under post-1976 rules, fixation in a tangible medium, not publication, is
the trigger for copyright. See supra text accompanying note 10.
SPRIGMAN FINAL 12/17/2004 3:36 PM
relatively recent instance is the song “Superman,” a hit in 1986 for the Athens,
Georgia, band R.E.M. “Superman” had originally been released in 1969 by The
Clique, an obscure Houston, Texas, group. The Clique’s version of the song
was not a hit (in fact, it was the “B side” to The Clique’s only hit single, “Sugar
On Sunday”), and the band released only one album which, by the time R.E.M.
recorded its cover, had long been “out of print.” Due, however, to the
popularity of the R.E.M. cover version of “Superman,” a compilation recording
of The Clique’s work was reissued in 1998.105
Reworkings of musical performances are addressed by the provision of the
Copyright Act imposing compulsory licenses for “mechanical rights”—i.e., an
automatic license that gives artists the ability to record and distribute their own
versions of musical compositions for a fee set by statute.106 But for derivative
works other than new performances of musical compositions, an author
wishing to use even the most obscure and commercially valueless material
must identify a rightsholder and ask permission. The necessity of identifying
rightsholders and negotiating rights raises the cost of creating derivative works.
Consequently, output of potentially valuable derivative works will fall under an
unconditional copyright regiume, in comparison to a conditional regime in
which commercially valueless source materials are filtered out of the copyright
system at their inception.
Although the utility of the registration and notice formalities seems
obvious, they have more often been viewed, on balance, as a hindrance. A 1904
report by the Register of Copyrights makes that point, lamenting that “a system
has gradually grown up under which valuable literary rights have come to
depend upon exact compliance with the statutory formalities which have no
relation to the equitable rights involved, and the question may very well be
raised whether this condition should be continued.”107 Criticism of formalities
and tales of hardship arising from accidental noncompliance abound in the
historical copyright literature.108
use without the right to make derivative works,118 and use with the requirement
that the user make freely usable any derivative work created using the original
source material (referred to as the “share-alike” license,119 similar in purpose to
the “copyleft” movement’s Open Software License120 and GNU General Public
License121). Creative Commons also provides a “no rights reserved” public
domain dedication license, which provides a perpetual and unconditional
license “for the benefit of the public at large and to the detriment of the
Dedicator’s heirs and successors,”122 as well as a “Founders’ Copyright”
license, under which rightsholders agree to dedicate their work to the public
domain after either a fourteen- or twenty-eight-year period.123 I will return to
Creative Commons later as we consider how to build and implement new-style
formalities.
2. Renewal
Until it was eliminated by the 1976 Act (for pre-1978 works) and the
Copyright Renewal Act (for all other works), the renewal formality served as
another filter, one that operated later in the lifecycle of the copyrighted work as
an ex post test of commercial viability.
The effect of the renewal requirement was, again, to measure authors’
desire for protection. The mechanism was the same as that for the initial filter
of registration, only it measured not whether a work had commercial value at
its inception, but whether its value was enduring. Works that retained
commercial value at the end of the initial copyright term (first fourteen and
later twenty-eight years) were renewed. Authors would not bother, however, to
renew works that ceased to profit them at the end of the initial term and for
which they held no realistic expectation of future profit. Historically,
approximately 15% of works were renewed, meaning that 85% of works moved
into the public domain—by consent of rightsholders—after a relatively short
term of protection.
Copyright Office data on renewal rates suggests that many authors place a
low value on continued copyright protection at the end of an initial copyright
term. In a 1961 report based on data subsequent to the 1909 Act, the head of the
Copyright Office’s Examining Division stated that, for the minority of
published works that were registered and for which notice of copyright was
given, less than 15% of all copyrights were being renewed.124
Landes and Posner, looking at Copyright Office data from 1910 to 2000,
arrived at the same average rate of renewal—around 15%.125 Figures 2 and 3,
above, graph the annual number of renewals and the rate of copyright renewal,
respectively, during that period. The data shows that the gross number of
renewals grew by a factor of more than ten between 1910 and 1991. But the
percentage of works renewed remained low throughout the period. Working
from the same data used by the Copyright Office, Landes and Posner estimate
the rate of renewal at less than 11% for the period between 1883 and 1964,
even though the renewal fee was trivial throughout this period.126 The rate of
renewal rose somewhat between 1980 and 1990, when it reached its single-year
high of 22%. Beginning in 1992, however, the rate began to decline sharply. As
noted above,127 in 1992, renewal for works copyrighted between 1964 and
1977 became automatic under the Copyright Renewal Act, and the decline in
renewal was in large part driven by the elimination of renewal as a formality.
But because the 1992 amendments did not eliminate all incentive to renew a
work, Landes and Posner argue that the decline in renewal is also likely to be
related to increases in the renewal fee, which doubled to $12 in 1991, rose to
$20 in 1993, and rose again to $45 in 2000.
If the general rate of renewal is low, renewal rates of certain significant
classes of works were even lower: the renewal rate for books has averaged less
than 8%, and for graphic arts approximately 3%.128 The average renewal rate
over this period for music is higher (32%),129 which one would expect given
the regularity with which even very old songs are reworked with new
performers and arrangements. But the renewal rate even for music peaked in
1956 and fell steadily after that. By 1969, the end of the data period for
renewals disaggregated by type of work, the renewal rate for music had fallen
almost to the historical norm for all works of around 15%.130
In sum, the renewal data reinforces what the registration data suggests—
that the difference between an unconditional and a conditional copyright
system, in terms of the scope of the works each system reaches, is profound. In
a conditional system, a substantial fraction of copyrightable works is not
valuable enough at inception to merit the investment necessary to secure
protection. And only a small portion of works retains enough value at the end
of an initial term to merit renewal. Using Copyright Office registration and
renewal data, Landes and Posner estimate an average annual depreciation rate
for copyrighted works ranging between 5.4% and 12.2%,131 which results in an
average expected commercial life for copyrighted works ranging from 8.2 to
18.5 years.132 Working from copyrights registered in a single year, 1934, the
authors estimate that 50% of the registered works had fully depreciated in just
10 years, 90% in 43 years, and 99% in 65 years.133 These findings are
supported by the results of a 1998 study by the Congressional Research Service
(CRS), which examined a sample of copyrights renewed after an initial term of
twenty-eight years. The CRS study concluded that only 11% of renewed
copyrights in books, 12% in musical works, and 26% in motion pictures had
some continuing commercial value.134
The renewal formality made the “real” term of copyright (in contrast to the
nominal term set out in the copyright statutes) very short by our current
standards. For the subset of works that was not eliminated by the initial
(registration and notice) filter and was therefore subject to copyright, the
longest effective copyright term prior to the 1976 Act, at an average renewal
rate of 15%, was 32.2 years.135 Using the highest historical renewal figure for
all works, 22% (in 1990), the average term of copyright would be 34.2 years.136
Under our current unconditional copyright system, there is no longer any
filtering mechanism tailoring the terms of individual works, and, consequently,
the real and nominal copyright terms have converged. The 1976 Act switched
from a fixed term of years to an indeterminate term for works by individual
authors: at first, life of the author plus fifty years, later extended by the CTEA
to life plus seventy years. For corporate works—known in the argot as “works
for hire”—and anonymous works, the 1976 Act fixed a term of seventy-five
years from the date of publication or one hundred years from creation,
whichever expired first. The CTEA extended those terms to ninety-five and one
hundred twenty years, respectively.
The copyright term is now sufficiently long that the net present value to the
rightsholder of a copyright is practically indistinguishable from what it would
be under a perpetual term. In an amicus curiae brief submitted to the Supreme
Court in support of the petitioners in Eldred v. Ashcroft, a group of economists
that included Nobel Prize winners George Akerlof, Kenneth Arrow, James
Buchanan, Ronald Coase, and Milton Friedman argued that the current, post-
CTEA copyright term of life plus seventy years has a net present value that is
99.88% of the value of a perpetual term.137
That the copyright term is now effectively perpetual is an odd development
in a country whose constitution specifies that copyrights may be granted only
for “limited times.”138 As will be discussed in greater detail later in this Article,
The social cost of monopoly. Any copyright system that grants exclusive
rights, whether based in a utilitarian or natural/moral rights conception,
imposes a number of different social costs. First, there is an obvious economic
cost, which is a specific instance of the general problem of monopoly: If a
particular creative work has a market value, exclusive rights will enable the
creator to charge a supracompetitive price. Consequently, access to the work
will be denied to those who value it in excess of the competitive price, but less
than the supracompetitive price that the monopolist is able to command.141
Copyright, then, creates deadweight losses in markets for expression.
The monopoly costs of copyright, while very real for works that possess
significant commercial value, are relatively unimportant to the commercially
valueless or exhausted works that conditional copyright filtered out but
unconditional copyright locks up. While an unconditional system keeps
economically spent works under copyright, the persistence of exclusive rights
can do little to raise the price of a piece of expression that is commercially
valueless—the rightsholder, in such an instance, may have a notional
“monopoly” but lack any power to demand a supracompetitive price.
Copyright’s burdens on speech. Much more important for our purposes are
the two types of “cultural” costs imposed by copyright. Copyright imposes a
First Amendment cost inhering in the restrictions on free speech imposed when
rightsholders are allowed to prevent copying of their works. The recent
copyright dispute involving Diebold Election Systems provides an example of a
potentially significant First Amendment cost imposed by copyright.
Diebold manufactures electronic voting machines. Sometime in early 2003
a hacker broke into the company’s computer systems and stole a large number
of internal e-mails and memoranda. Some of the stolen documents included
discussions of software bugs in Diebold voting machines and warnings that the
machines may produce unverifiable results and are poorly protected against
hackers.142 In August 2003, an unknown person mailed approximately 13,000
pages of the stolen data to a number of activists concerned with electronic
voting, many of whom published the Diebold e-mails and memos, or linked to
those documents, on their websites.
In response, Diebold sent dozens of cease-and-desist notices, pursuant to
the “notice and take-down” provisions of the Digital Millenium Copyright Act
(DMCA),143 to website publishers and Internet service providers (ISPs),
demanding that they remove the documents from websites and cease linking to
141. For a more fully developed account of the benefits and costs of copyright, see
Linda R. Cohen & Roger G. Noll, Intellectual Property, Antitrust and the New Economy, 62
U. PITT. L. REV. 453 (2001).
142. For factual background, see First Amended Complaint, Online Policy Group v.
Diebold Inc., No. C 03-04913 JF, 2004 WL 2203382 (N.D. Cal. Sept. 30, 2004), available at
http://www.eff.org/legal/ISP_liability/OPG_v_Diebold/First_Amended_Complaint.pdf (last
visited Sept. 9, 2004).
143. Pub. L. No. 105-304, 112 Stat. 2860 (1998). The DMCA provides a “safe harbor”
provision as an incentive for ISPs to take down user-posted content when they receive cease-
and-desist letters such as the ones sent by Diebold. 17 U.S.C. § 512(c) (2000). By removing
the content, or forcing the user to do so, for a minimum of ten days, an ISP can immunize
itself from any copyright claim. 17 U.S.C. § 512(g)(2)(C).
SPRIGMAN FINAL 12/17/2004 3:36 PM
144. Online Policy Group v. Diebold Inc., No. C 03-04913 JF, 2004 WL 2203382
(N.D. Cal. Sept. 30, 2004), available at http://cyberlaw.stanford.edu/about/cases/20040930_
Diebold_SJ_Order.pdf (last visited Nov. 11, 2004).
145. Note that under the pre-1976 regime, where federal copyright commenced upon
“publication” and perpetual common law copyright applied to unpublished works, Diebold’s
unpublished corporate memoranda would likely have been subject to common law copyright,
and Diebold would still have had a powerful copyright lever. (I say “likely,” rather than
“certainly,” because the pre-1976 rules determining what constituted “publication”—the
trigger that terminated state common law copyright and moved a work into the federal
system—were the subject of substantial debate and confusion. See William S. Strauss,
Protection of Unpublished Works, in 1 STUDIES ON COPYRIGHT, supra note 124, at 189, 198-
205. In contrast, under a reformalized version of our post-1976 system, where federal
copyright arises upon fixation, documents like the Diebold memoranda would seldom enter
the copyright system.
146. See, e.g., Religious Tech. Ctr. v. Netcom On-Line Communication Servs., Inc.,
923 F. Supp. 1231 (N.D. Cal. 1995) (holding that copyright could be used to prevent
distribution of Church of Scientology materials). See generally Jed Rubenfeld, The Freedom
SPRIGMAN FINAL 12/17/2004 3:36 PM
copyrighted works that may be required to make a fair use.149 That holding
threatens to make fair use a mirage as technology shifts creativity from analog
to digital media.
Assessing whether fair use includes some notion of fair access is one of the
areas where copyright analogies from the analog world break down. Questions
of access in the analog world only come up when the access sought is to
someone else’s property. How could it be otherwise, since possession of analog
materials necessarily entails access to make a copy? In contrast, the issue we
confront in the digital world is access to one’s own property for the purpose of
making fair uses. Put that way, a judicially established right of access would
not seem to require much of a logical leap. But to say that there is not currently
a judicially established right of access to make fair uses of copyrighted
materials is not especially meaningful. In the analog world we lived in until
only yesterday (at least in the time frame of the law), the question simply never
came up.
For the moment, however, it is clear that if fair use in the digital
environment depends on fair access, the opportunities to make fair uses in a
world of strong (and legally enforced) encryption are going to be substantially
restricted. As with the idea/expression dichotomy, the shift from the analog to
the digital environment has altered (or, more precisely, has constricted) the
“traditional contours” of the fair use doctrine. In the case of fair use, however,
the effect is worse, because it is not simply the product of shifting technologies,
but of government action—i.e., the DMCA, which prohibits circumvention
without providing any exception for fair use access.
So if the idea/expression dichotomy and fair use have been enfeebled, are
there other “traditional contours” of copyright that remain vigorous enough to
mediate between copyright and the First Amendment? There is—or at least
was—a third “buffer” that played a very significant role: copyright formalities.
Under conditional copyright, formalities served to limit copyright protection to
works that had independent value as expression. Works that lacked expression
value ordinarily would not be copyrighted. The purpose of copyright is to
incent expression ex ante, not to serve as a locking mechanism ex post.
Copyright formalities created an incentive structure that aligned the material
protected under copyright with the overarching justification for the regime.
With the disappearance of formalities, perversions of copyright like we observe
in the Diebold dispute become not only possible, but inevitable.
Copyright’s burden on creativity. In addition to burdening free speech,
copyright also imposes costs on future creativity by shrinking the stock of
preexisting materials available to future creators for use as building blocks in
new works, which reduces consequentially the production of new works.150
149. See Universal City Studios, Inc. v. Corley, 273 F.3d 429, 458-59 (2d Cir. 2001).
150. See LANDES & POSNER, supra note 1, at 58-60 (providing examples of
SPRIGMAN FINAL 12/17/2004 3:36 PM
transformative use of preexisting materials in the works of Shakespeare, Yeats, and Eliot,
among others).
151. See 2 GOLDSTEIN, supra note 140, § 7.2.1.2.
152. Id. § 7.4.1.1.
SPRIGMAN FINAL 12/17/2004 3:36 PM
Why did the “traditional contours” of pre-1976 U.S. copyright law require
compliance with so many bothersome formalities? In order to understand the
role of formalities in our pre-1976 conditional copyright system, it is helpful
first to look at the source of Congress’s authority to enact copyright laws: the
Constitution’s Intellectual Property Clause. The clause does not itself require
that Congress install any particular formality in copyright laws. Yet it reflects
an original understanding of the purpose of copyright that led, in the early
copyright statutes and for almost two centuries thereafter, to a system that
relied heavily on formalities.
Article I, Section 8, Clause 8 of the U.S. Constitution confers upon
Congress authority “[t]o promote the Progress of Science and useful Arts, by
securing for limited Times to Authors and Inventors the exclusive Right to their
respective Writings and Discoveries.” As the Supreme Court has recognized,
the Intellectual Property Clause is “both a grant of power and a limitation.”153
But aside from formulating that aphorism, the Supreme Court has done little to
limn either the power or the limitation, or to define the judicial role in ensuring
that Congress’s copyright lawmaking remains within the boundaries set out by
the clause.
The most significant continuing dispute in the interpretation of the
Intellectual Property Clause involves the most basic question of interpretation:
which part of the clause sets out the enumerated power? Edward Walterscheid
has argued that the grant of power resides in the “promote . . . Progress”
language, and that that power, moreover, is a general one that authorizes
Congress to undertake a variety of schemes, such as the funding of medical
research or grants to arts organizations, with the common purposes of
encouraging discovery and spreading culture.154 Walterscheid contends that the
second part of the clause (the “exclusive Right[s]” language) was added only
for the purpose of making clear that, subject to certain limitations, Congress
was authorized to grant patents and copyrights as part of its general power to
advance learning.155
More often, however, it has been argued that the power resides in the
“exclusive Right[s]” part of the clause. According to this interpretation, the
power granted is specific—i.e., Congress is authorized to grant limited-time
exclusive rights for the purpose of advancing learning. The federal
156. See, e.g., DAVID P. CURRIE, THE CONSTITUTION IN CONGRESS: THE FEDERALIST
PERIOD, 1789-1801, at 93 (1997) (arguing that the Intellectual Property Clause confers “not a
general power to ‘promote the progress of science and the useful arts,’ but only the power to
grant limited exclusive rights in order to accomplish that goal”).
157. 537 U.S. 186, 212 (2003) (quoting Graham, 383 U.S. at 6).
158. Compare Schnapper v. Foley, 667 F.2d 102, 112 (D.C. Cir. 1981) (concluding
that introductory language does not limit congressional power), and Mitchell Bros. Film
Group v. Cinema Adult Theater, 604 F.2d 852, 860 (5th Cir. 1979) (same), and 1 NIMMER &
NIMMER, supra note 13, § 1.03 (arguing that the promote progress phrase “is in the main
explanatory of the purpose of copyright, without in itself constituting a rigid standard against
which any copyright act must be measured”), with Twentieth Century Music Corp. v. Aiken,
422 U.S. 151, 156 (1975) (suggesting, in dicta, that the promote progress language may
inform the meaning of otherwise ambiguous statutory language: “When technological
change has rendered its literal terms ambiguous, the Copyright Act must be construed in
light of [its] basic purpose.”).
159. See Graham, 383 U.S. at 5-6 (“Congress in the exercise of the patent power may
not overreach the restraints imposed by the stated constitutional purpose.”); Great Atl. &
Pac. Tea Co. v. Supermarket Equip. Corp., 340 U.S. 147, 154 (1950) (Douglas, J.,
concurring) (“Congress acts under the restraint imposed by the statement of purpose in Art.
I, § 8.”).
SPRIGMAN FINAL 12/17/2004 3:36 PM
the Court stated, refer to inventions (i.e., patents) as well as literary works, and
“it has never been pretended, by any one, either in this country or in England,
that an inventor has a perpetual right, at common law, to sell the thing
invented.”168 Neither the clause nor the 1790 Act provided any reason to
distinguish between the source of the exclusive right for inventions versus
literary works. The Court also found, in a passage notable more for its
forcefulness than its logic, that the language of the 1790 Act established that
Congress was creating a right, not sanctioning an existing one:
That congress, in passing the act of 1790, did not legislate in reference to
existing rights, appears clear, from the provision that the author, &c ‘shall
have the sole right and liberty of printing,’ &c. Now if this exclusive right
existed at common law, and congress were about to adopt legislative
provisions for its protection, would they have used this language? Could they
have deemed it necessary to vest a right already vested. Such a presumption is
refuted by the words above quoted, and their force is not lessened by any other
part of the act.169
Having found that copyright in published works was a right created by
statute, the Court held that noncompliance with the statutory prerequisites—
including those, such as deposit and notice, performed subsequent to
publication—vitiates the copyright:
[T]he inquiry is made, shall the non performance of these subsequent
conditions operate as a forfeiture of the right?
The answer is, that this is not a technical grant of precedent and subsequent
conditions. All the conditions are important; the law requires them to be
performed; and, consequently, their performance is essential to a perfect
title.170
Facing an incomplete factual record, the Court remanded to the circuit court for
a determination whether Wheaton had complied with the deposit and notice
formalities.171
If the Supreme Court in Wheaton made clear that copyright is established
by law, rather than merely enforced by it, then the obvious question is “For
what purpose has the right been established?” Uniquely among the legislative
powers enumerated in Article I, Section 8, the Intellectual Property Clause ties
the power to grant patents and copyrights to a specified purpose—the
promotion of progress in “science” (by which the Framers meant all forms of
knowledge, including literature and the arts) and the “useful arts” (by which the
Framers meant patentable inventions). The justification for copyright (and
patent) set out in the clause is utilitarian: Congress is authorized to create
exclusive rights not as an end in itself, but merely as a means of “promoting
172. Abraham Lincoln, Second Lectures on Discoveries and Inventions (Feb. 11,
1859), in 3 THE COLLECTED WORKS OF ABRAHAM LINCOLN 356, 363 (Roy P. Basler ed.,
1953).
173. Stewart E. Sterk, Rhetoric and Reality in Copyright Law, 94 MICH. L. REV. 1197,
1203 (1996); cf. Mazer v. Stein, 347 U.S. 201, 219 (1954) (“Sacrificial days devoted to such
creative activities deserve rewards commensurate with the services rendered.”).
174. 334 U.S. 131, 158 (1948).
175. 347 U.S. at 219.
176. 464 U.S. 417, 429 (1984); see also Twentieth Century Music Corp. v. Aiken, 422
U.S. 151, 156 (1975) (“The immediate effect of our copyright law is to secure a fair return
for an ‘author’s’ creative labor. But the ultimate aim is, by this incentive, to stimulate artistic
creativity for the general public good.”).
SPRIGMAN FINAL 12/17/2004 3:36 PM
If our copyright system were driven purely by the utilitarian concerns that
undergird the Intellectual Property Clause, rather than by more concrete
political considerations,178 we might expect our copyright term to be relatively
short and formalities to permeate the law. But while Congress and the courts
have paid lip service to utilitarian copyright, they have, on a practical level,
acquiesced to developments in the law—including the move from conditional
to unconditional copyright, the broadening of the rights granted by copyright to
cover nearly every conceivable use of the protected work (including the
production of derivative works), and the extension of the term to a point that
the return to rightsholders is indistinguishable from that produced by perpetual
copyright—that together have made our theoretically utilitarian system almost
177. H.R. REP. NO. 60-2222, at 7 (1909). Almost a century earlier, Thomas Jefferson
expressed the same idea with characteristic felicity:
If nature has made any one thing less susceptible than all others of exclusive property, it is
the action of the thinking power called an idea, which an individual may exclusively possess
as long as he keeps it to himself; but the moment it is divulged, it forces itself into the
possession of every one, and the receiver cannot dispossess himself of it. Its peculiar
character, too, is that no one possesses the less, because every other possesses the whole of it.
He who receives an idea from me, receives instruction himself without lessening mine; as he
who lights his taper at mine, receives light without darkening me. . . . Inventions then cannot,
in nature, be a subject of property. Society may give an exclusive right to the profits arising
from them, as an encouragement to men to pursue ideas which may produce utility, but this
may or may not be done, according to the will and convenience of the society, without claim
or complaint from anybody.
Letter from Thomas Jefferson to Isaac McPherson (Aug. 13, 1813), in 13 THE WRITINGS OF
THOMAS JEFFERSON, supra note 138, at 326, 333-34.
178. See generally LANDES & POSNER, supra note 1, ch. 15 (discussing the political
economy of intellectual property law).
SPRIGMAN FINAL 12/17/2004 3:36 PM
term for the large majority of works was twenty-eight years; today copyright
imposes a uniform term lasting, on average, three times as long. Considering
the distance our law has traveled in the evolution from conditional to
unconditional copyright, it is difficult at this point to understand which of
copyright’s “traditional contours” the Court believes remain undisturbed.
It is not enough to look, as the Court did in Eldred, at the copyright term in
isolation. When one looks more closely at the effects of a series of seemingly
minor changes to the copyright law—changes that are unrelated (or at least not
facially related) to the copyright term—it is clear that the “traditional contours”
of our copyright system went through a disjunction during the move from
conditional to unconditional copyright. Of course, it may be difficult just now,
considering the result in Eldred, to see exactly how the increasing detachment
of U.S. copyright law from its constitutional underpinnings could lead to
judicial intervention and invalidation of elements of the law. The following are
two possible arguments that have been advanced in Kahle v. Ashcroft, a lawsuit
filed recently in a federal court in California.184
Copyright’s potential collision with the Constitution could take the form of
an Intellectual Property Clause challenge to copyright extension that is
somewhat narrower—but no less potentially disruptive to the status quo—than
that posed in Eldred. The argument urges a reevaluation of the historical
record, based on an observation about the effect of the renewal formality on
copyright extensions that was never raised in Eldred.
While Congress had extended the term of subsisting copyrights on several
occasions prior to the CTEA, in every case before the CTEA, the subsisting
copyrights whose terms were extended were required to pass at some point
through the filter of renewal. The 1831 Act extended the initial term of
subsisting copyrights from fourteen to twenty-eight years, but within a regime
that required copyright owners to renew their copyright to secure the benefits of
the maximum term of forty-two years.185 The 1909 Act likewise extended the
renewal term of subsisting copyrights, but the Act expressly limited its effect to
works that had been renewed.186 Even the 1976 Act, which began the march
toward unconditional copyright and again extended the term of subsisting
copyrights, limited its extension to works that had been renewed.187
184. Kahle Complaint, supra note 5. The author is co-counsel for plaintiffs in this
lawsuit.
185. See supra notes 33, 35 and accompanying text.
186. See supra text accompanying notes 38-42.
187. See 1976 Act, supra note 6, § 304.
SPRIGMAN FINAL 12/17/2004 3:36 PM
works would never have had their copyrights renewed, the law automatically
extended their terms by sixty-seven years. This is the first category of
copyrighted works in U.S. history which has had its term extended
automatically without ever passing through the filter of renewal.
Because these changes have altered the “traditional contours” of copyright,
they should be evaluated under heightened First Amendment scrutiny. But even
under the less exacting rational basis standard, the burdens created by these
changes for certain categories of copyrighted works far outweigh any plausible
benefits.
The Berne Convention, which dates from 1886, was the fruit of
negotiations that had been proceeding since the first International Congress of
Authors and Artists met in Brussels in 1858.194 In its current form,195 the Berne
Convention obliges signatories to honor two basic principles: (1) a “national
treatment” principle reqiring all signatory nations to grant the same rights to
foreign authors that they grant to their own authors; and (2) a “baseline
protection” principle requiring signatory countries to adhere in their domestic
193. See, e.g., Copyright Law Revision, supra note 108, at 68 (statement of Abraham
L. Kaminstein, Register of Copyrights).
194. See Jane C. Ginsburg, International Copyright: From a “Bundle” of National
Copyright Laws to a Supranational Code?, 47 J. COPYRIGHT SOC’Y U.S.A. 265, 267 (2000).
195. The Berne Convention has been revised repeatedly; particular revisions are
referred to as “acts.” The group of countries that are signatories to the Convention, referred
to collectively as the “Berne Union,” has an existence separate from any particular act. When
the Convention is revised, Union members are not required to adhere to the new revision as a
condition to remaining within the Union. A country may join the Union at any time by
acceding to the most recent version of the Convention. The treaty obligations of any
particular Union member are measured by the terms of the particular act or acts to which that
member has acceded. See PAUL GOLDSTEIN, INTERNATIONAL COPYRIGHT 20-21 (2001).
SPRIGMAN FINAL 12/17/2004 3:36 PM
196. Berne Convention, supra note 14, art. 5(1). Under the Berne Convention’s “points
of attachment” rules, a work is entitled to Berne Convention protection in signatory nations
if its author is a national or domiciliary of a signatory state or if the work is first or
simultaneously published in a signatory state. See id. arts. 3(1)(a), 3(1)(b), 3(2).
197. Id. art. 7(1).
198. Id. art. 5(2).
199. Proclamation No. 3, 27 Stat. 981 (July 1, 1891).
200. Id.
201. Proclamation No. 24, 27 Stat. 1021 (Apr. 15, 1892).
202. Universal Copyright Convention, Geneva Text, Sept. 6, 1952, art. 3(2), 6 U.S.T.
2731, 2735, 216 U.N.T.S. 132, 136.
203. Graeme W. Austin, Does the Copyright Clause Mandate Isolationism?, 26
COLUM.-VLA J.L. & ARTS 17, 42 (2002).
204. Id. This “back door” to Berne has, since 1914, been subject to the power of Union
members to retaliate against authors who are nationals of non-Union countries but obtain
Berne protection through first publication in a Union country, if the author’s country of
nationality “fails to protect in an adequate manner the works of authors who are nationals of
one of the countries of the Union.” Berne Convention, supra note 14, art. 6(1).
SPRIGMAN FINAL 12/17/2004 3:36 PM
205. Berne Convention for the Protection of Literary and Artistic Works, Sept. 9, 1886,
art. 2(2), reprinted in ARPAD BOGSCH, BERNE CONVENTION FOR THE PROTECTION OF
LITERARY AND ARTISTIC WORKS, FROM 1886 TO 1986, at 228, 228 (World Intellectual Prop.
Org. Public’n No. 877(E), 1986) [hereinafter BERNE CONVENTION FROM 1886 TO 1896].
206. Berne Convention for the Protection of Literary and Artistic Works, Berlin Act,
Nov. 13, 1908, art. 4(2), reprinted in BERNE CONVENTION FROM 1886 TO 1986, supra note
205, at 229, 229.
207. Berne Convention, supra note 14, art. 5(2).
208. WORLD INTELLECTUAL PROP. ORG., GUIDE TO THE BERNE CONVENTION FOR THE
PROTECTION OF LITERARY AND ARTISTIC WORKS (PARIS ACT, 1971) 33 (1978) [hereinafter
WIPO GUIDE]; see also SAM RICKETSON, THE BERNE CONVENTION FOR THE PROTECTION OF
LITERARY AND ARTISTIC WORKS: 1886-1986, at 222-23 (1987).
209. See supra text accompanying notes 43-54.
210. See STEPHEN M. STEWART, INTERNATIONAL COPYRIGHT AND NEIGHBOURING
RIGHTS 106 (1983) (“[T]he necessity to register before bringing an action would probably be
regarded as a ‘formality’ as it negates the ‘exercise’ of the right without such registration.”);
Mayer Gabay, The United States Copyright System and the Berne Convention, 26 BULL.
SPRIGMAN FINAL 12/17/2004 3:36 PM
COPYRIGHT SOC’Y U.S.A. 202, 208 (1979) (“It is true that registration is ‘permissive’ and
does not, under the 1976 Act, constitute a condition precedent for acquisition of copyright.
But these factors merely give rise to a bare right that is incapable of being exercised in a
U.S. court of law until registration is effected.”).
211. The fact that registration affords successful infringement plaintiffs the opportunity
to collect statutory damages and attorney’s fees, for example, is not believed to violate the
Berne Convention, because that instrument does not itself require that a country provide for
such recoveries. See Gabay, supra note 210, at 209-10; Melville B. Nimmer, Implications of
the Prospective Revisions of the Berne Convention and the United States Copyright Law, 19
STAN. L. REV. 499, 514 (1967).
212. Berne Convention, supra note 14, art. 5(2).
213. WIPO GUIDE, supra note 208, at 33. WIPO, one of the sixteen specialized
agencies of the United Nations system of organizations, is the principal forum for
negotiation of international intellectual property agreements. Headquartered in Geneva,
Switzerland, WIPO administers twenty-three international treaties dealing with different
aspects of intellectual property protection. WIPO counts 179 nations as member states. For
more information on WIPO’s mission, see World Intellectual Prop. Org., Medium-Term
Plan for WIPO Program Activities, at http://www.wipo.int/about-wipo/en/dgo/pub487.htm
(last visited Oct. 31, 2004).
214. See supra note 81.
215. Berne Convention, supra note 14, art. 7(1); see also id. art. 7(2) (establishing for
cinematographic works a minimum term of fifty years after publication, or, for unreleased
films, fifty years after production); id. art. 7(3) (establishing for anonymous and
pseudonymous works a minimum term of fifty years after publication); id. art. 7(4)
(establishing for photographs and works of applied art a minimum term of twenty-five years
following a work’s production in countries that protect these types of works).
SPRIGMAN FINAL 12/17/2004 3:36 PM
216. Alan Story, Burn Berne: Why the Leading International Copyright Convention
Must Be Repealed, 40 HOUS. L. REV. 763, 771 (2003); see also Hesse, supra note 139, at 40
(“[Berne] tended to strengthen universalist claims for protection of inviolable natural rights
against statutory limits imposed by particular nations on utilitarian grounds.”).
217. The European Union required member nations to adopt a uniform life-plus-
seventy term in a 1996 directive. Council Directive 93/98/EEC of 29 October 1993
Harmonizing the Term of Protection of Copyright and Certain Related Rights, 1993 O.J. (L
290) 9 [hereinafter EU Directive]. The EU Directive has now become the global benchmark.
It was cited as a factor in the CTEA’s installment of an identical term in U.S. law, see Eldred
v. Ashcroft, 537 U.S. 186, 195-96 (2003), and is reflected in domestic legislation in a
number of countries both within and outside Europe, see Shauna C. Bryce, Life Plus
Seventy: The Extension of Copyright Terms in the European Union and Proposed
Legislation in the United States, 37 HARV. INT’L L.J. 525, 529 (1996).
SPRIGMAN FINAL 12/17/2004 3:36 PM
familiarity, and then to obtain and fill out forms in a variety of languages—
would be difficult, expensive, and often result in unintentional noncompliance
and the loss of valuable rights. Article 5(2) of the current Paris Text of the
Berne Convention was promulgated in 1971, but the nature of the problem had
not changed in the intervening sixty-three years: copyright systems remained
substantively and procedurally diverse, the mechanisms of compliance in many
countries remained balky, and the costs of informing oneself about
requirements in different countries, and then complying with them, remained
high.
Since 1971, however, there has been a series of technological changes that
could make compliance with a redesigned set of formalities quick and easy.
Those changes involve, of course, computers and the Internet. Now it is
possible for an author publishing a work internationally to comply with
formalities in his or her home country, or in the country of a work’s first
publication, and to have the data generated by that compliance formatted and
transmitted reliably and nearly costlessly to other jurisdictions. But changes in
technology alone are not enough—changes to the law are also required. To
make compliance cheap, the law must ensure that data generated in one
jurisdiction will be sufficient to permit compliance in any jurisdiction that
chooses to reintroduce formalities into its domestic copyright laws.
The simplest way to take advantage of what technology now allows would
be to propose a new Berne text that removes the prohibition in Article 5(2) of
the current Paris Act and replaces it with a provision allowing member
countries to impose formalities, provided that they adhere to a set of standards
that make formalities “interoperable” across jurisdictions.227 What would the
changes to the Berne Convention look like?
227. The North American Free Trade Agreement (NAFTA) also contains provisions
related to intellectual property. See, e.g., North American Free Trade Agreement, Dec. 17,
1992, U.S.-Can.-Mex., arts. 1708 (trademark), 1709 (patent), 1710 (semiconductor design),
1711 (trade secrets), 32 I.L.M. 605, 672-75 (implemented by 107 Stat. 2057). In particular,
NAFTA Article 1703(2) mirrors the Berne Article 5(2) proscription of formalities, and must
likewise be modified to permit reformalization of domestic law in the United States. NAFTA
Article 1703(2) provides that “[n]o Party may, as a condition of according national treatment
under this Article, require rights holders to comply with any formalities or conditions in
order to acquire rights in respect of copyright and related rights.” Id. art. 1703(2). Since
NAFTA signatories Canada, Mexico, and the United States are all parties to a text of the
Berne Convention that prohibits formalities, NAFTA Article 1703(2) is largely duplicative
of Berne on the issue of formalities.
SPRIGMAN FINAL 12/17/2004 3:36 PM
For countries that, unlike the United States, do not include protection for
performances, sound recordings (also referred to as “phonograms”), and
broadcasts in their copyright laws, but locate them instead in separate
“neighboring rights” statutes, reformalization of domestic law must include
changes to the law governing both types of rights. Similarly, the same
reciprocity principle that would be installed into the Berne Convention must
also be introduced into the applicable international agreements governing
neighboring rights—an issue of some complexity.
SPRIGMAN FINAL 12/17/2004 3:36 PM
examine how such a system would work with respect to each of the familiar
types of formalities.
232. Indeed, because notice on existing works usually cannot be updated, notice can
give misleading information regarding current ownership. For countries that choose to
require notice, the marking requirement should be refocused to provide would-be users with
information facilitating access to the registry, where the most up-to-date information on
ownership would be available.
SPRIGMAN FINAL 12/17/2004 3:36 PM
minimum copyright term of life of the author plus fifty years.233 The answer is
likely no. Berne’s prohibition of formalities is contained in an article separate
from its minimum term requirements. Remove the current ban on formalities,
and nothing in the Convention specifies that the term, if offered equally to
every author, must be enjoyed in full by every author without condition.
The second problem is a more practical one: how to coordinate renewal
across jurisdictions when different Berne member states may impose renewal
requirements at different points in the copyright term. A simple application of
the reciprocity principle threatens to create substantial confusion. If a
rightsholder who complies with the renewal formality in the jurisdiction in
which a work was first registered is deemed to have complied with renewal in
any jurisdiction in which a renewal is required, then absent detailed knowledge
of the point at which renewal may be required in a potentially large number of
Berne jurisdictions, a would-be user will find it difficult to determine whether a
work has been timely renewed.
For example, suppose that a work is first registered in country A, which
imposes a renewal requirement at thirty years. Twenty-five years into the
work’s term, a would-be user in country B inquires whether the work is in the
public domain. Country B imposes a renewal formality at fifteen years. The
user sees that the work was registered twenty-five years ago; under country B’s
law, the work would have passed into the public domain when the rightsholder
failed to timely renew. But under country A’s laws, the work is still in its initial
term; renewal will not be required for another five years. Accordingly, under a
simple application of the reciprocity principle, unless the user understands (1)
where the work was first registered, (2) when the renewal requirement occurs
in that jurisdiction, and (3) that country A’s renewal requirement is the relevant
one, the user will not easily be able to determine whether the work is in the
public domain.
These information problems can be mitigated, of course, even if they
cannot be eliminated. The standardized registration and notice format should
include information on the nation of the work’s first registration, and that
information should be made available in all online registry sites maintained in
the various Berne nations. In addition, Berne signatories should be encouraged
to disseminate information about the rules governing renewal, and how to
determine which renewal term applies to a particular work. Taken together,
these measures might mean that the benefit, in terms of the simplicity of a
straightforward application of the reciprocity principle to renewal, outweighs
the cost in terms of the increased complexity of determining the status of rights.
An alternative, which would require a greater degree of coordination
among Berne nations, would be to standardize renewal terms for all
jurisdictions that reinstall a renewal formality in their domestic law. The
reciprocity principle would then apply to grant automatic compliance with all
Berne nations’ renewal requirements based on timely compliance with the
requirement in the country of first registration, or by renewing with WIPO, if
original registration was made with the WIPO registry. Based on the
depreciation calculations made by Landes and Posner, the Berne nations could
impose more than one renewal obligation during the copyright term. A first
renewal obligation set at ten years would move approximately 50% of
registered works into the public domain. A second renewal requirement set in
the vicinity of forty-three years would result in only 10% of the number of
originally registered works remaining under copyright. A third renewal
requirement set at sixty-five years would move all but 1% of the number of
originally registered works into the public domain. The works left under
copyright after sixty-five years would be those of truly enduring commercial
value for which the full term of copyright would be likely to provide significant
continuing benefits.
234. See Final Report, supra note 45, app. A at 622 (1986) (“[W]e have proposed what
we think are minimal amendments to the law, only where change is clearly required, based
upon widely shared understandings of Berne obligations . . . . A number of the alternatives
SPRIGMAN FINAL 12/17/2004 3:36 PM
Because the United States is a signatory to the UCC, and because before it
acceded to Berne the United States negotiated bilateral copyright agreements
with several nations that were not UCC signatories, it would be possible for the
United States to withdraw from Berne and rely instead on the UCC, which,
unlike Berne, allows the imposition of formalities for the works of both
domestic and foreign authors. This strategy would, however, impose
unacceptable costs, the largest of which would arise from our resulting
noncompliance with the TRIPs accord, which incorporates by reference
Berne’s standards, and with the North American Free Trade Agreement
(NAFTA), which replicates Berne’s ban on formalities. Although its
applicability is subject to considerable debate, it is also possible that a
provision of Berne’s Appendix Declaration would prevent U.S. authors from
claiming the benefits of the UCC in countries that are Berne signatories.235
A further cost of withdrawal would arise from Berne Article 6(1), which
permits Berne nations to restrict the protection accorded to works of authors
who are nationals of a non-Berne country that “fails to protect in an adequate
manner the works of [Berne nationals].”236 There is little commentary on this
provision, so it is difficult to forecast whether subjecting foreign works to
formalities (at least formalities that do not discriminate between domestic and
foreign works, and for which compliance is easy and cheap) would rise to the
level of a “fail[ure] to protect in an adequate manner” the rights of foreign
authors.
we propose for consideration seek to exploit the distinction in treatment between works of
foreign and national origin permitted by the Berne Convention.”).
235. RICKETSON, supra note 208, at 856.
236. Berne Convention, supra note 14, art. 6(1).
SPRIGMAN FINAL 12/17/2004 3:36 PM
237. LANDES & POSNER, supra note 1, ch. 8; William M. Landes & Richard A. Posner,
Indefinitely Renewable Copyright, 70 U. CHI. L. REV. 471 (2003).
238. Incompatibility with Berne’s rule against formalities could be avoided, of course,
by limiting the proposal to works of U.S. authors. But even such a limitation would not
prevent incompatibility with the life-plus-fifty minimum term requirement, which applies to
all works.
239. LANDES & POSNER, supra note 1, at 215 n.15.
240. In a forthcoming article, William Patry and Richard Posner argue that the
problems created by ever-longer copyright terms in a deformalized system should be
addressed by expanding the fair use doctrine to immunize use of works where a reasonable
inquiry fails to identify a rightsholder from whom a license may be sought. William F. Patry
& Richard A. Posner, Fair Use and Statutory Reform in the Wake of Eldred, 92 CAL. L. REV.
(forthcoming 2004) (working draft on file with author). The Patry and Posner article does
not attempt to describe in any detail what a “reasonable inquiry” is, or how reasonableness
SPRIGMAN FINAL 12/17/2004 3:36 PM
Another possible approach is set out in a bill currently before Congress, the
Public Domain Enhancement Act (PDEA).241 Sponsored by Representative
Zoe Lofgren, a California Democrat, the PDEA would give copyright owners
of works by U.S. authors unfettered rights for fifty years. At that point the
copyright holder would be required to file a notice of continuation and pay a $1
fee every ten years to continue the copyright. Because only a small number of
works would retain any commercial value at the expiry of the minimum term,
most copyright owners would not bother to file a notice of continuation and pay
the fee. On September 4, 2003, the PDEA was referred to the House
Subcommittee on Courts, the Internet, and Intellectual Property. There it has
languished.
Unlike the Posner and Landes proposal, the PDEA is very likely
compatible with Berne. The renewal requirement is limited to the works of U.S.
authors, thereby avoiding conflict with Berne’s rule against formalities.
Although the renewal provision may cut off a work’s copyright prior to the
expiration of Berne’s minimum term, that should not, for the reasons given
above, cause Berne noncompliance.242 Nonetheless, the PDEA is vulnerable to
the critique that its effect is limited to tinkering around the margins: A large
percentage of works are commercially valueless at inception or have an initial
value that is quickly depleted. All of these works, however, would continue
under the PDEA to be subject to a very long copyright term. While fifty years
is certainly better than life plus seventy years, it may reasonably be asked
whether the game is worth the candle.
5. New-style formalities
A fifth option, and by far the most attractive, is to formulate and install in
U.S. law a set of new-style formalities that apply to both domestic and foreign
works. New-style formalities would provide the filtering and information-
creation benefits of traditional formalities. However, there is a good argument
that, if structured properly, new-style formalities would not affect copyright’s
“enjoyment and exercise,” and would, therefore, comply with our Berne
obligations under the current Paris Act. This approach is attractive because it
would require changes only to U.S. law; Berne, TRIPs, and the other
international agreements that govern copyright and neighboring rights would
remain undisturbed. Integrating new-style formalities with the current text of
the Berne Convention does, however, raise several significant questions.
First, exactly what is a condition that interferes with the “enjoyment and
exercise” of copyright? A solid starting point is that that language at least
means that failure to comply with a formality cannot formally terminate the
right, or prevent it from arising in the first place. It is unclear how much further
“enjoyment and exercise” goes than that; I will return to this problem later.
Assuming for the moment that Article 5(2) allows a range of options short
of formal nullification of copyright, we are still faced with a difficult problem:
new-style formalities have to create a sufficient incentive for compliance to
construct a reliable record of ownership and to reliably signal copyright status,
but cannot use the forfeiture of rights to incent compliance.
The simplest solution would be to preserve formally voluntary registration,
notice, and recordation of transfers (and reestablish a formally voluntary
renewal formality) for all works, including works of foreign authors, but then
incent compliance by exposing the works of noncompliant rightsholders to a
“default” license that allows use for a predetermined fee. The royalty payable
under the default license would be low. Ideally, the royalty to license a work
that a rightsholder has failed to register, notice, reregister in the case of a
transfer (i.e., record), or renew should be set to approximate the cost of
complying with these formalities (i.e., the total cost of informing oneself about
the details of compliance and then satisfying them). That way a rightsholder
who expects his work to produce revenue exceeding the cost of complying with
the relevant formality will prefer to comply with the formality, whereas a
rightsholder who expects his work to produce revenue amounting to less than
the cost of compliance will prefer to expose his work to the default license.
(The rare rightsholder who estimates the likely revenues from his work to be
equal to the cost of complying with the formality will be indifferent between
compliance and exposure to the license.)
This system of formally voluntary formalities plus default licenses—which
I have referred to previously as “new-style” formalities—establishes indirectly
what the traditional system of compulsory formalities did directly: it eases
access to commercially valueless works for which protection (or the
continuation of protection) serves no purpose and focuses the system on those
works for which protection is needed to ensure that the rightsholder is able to
appropriate the commercial value of the expression. For the filtering function to
work, of course, the government would have to maintain an easily accessible
and up-to-date public registry. Given current computer database and search
technology, this would not be difficult.
SPRIGMAN FINAL 12/17/2004 3:36 PM
243. Ian Ayres & Robert Gertner, Filling Gaps in Incomplete Contracts: An Economic
Theory of Default Rules, 99 YALE L.J. 87, 97-98 (1989).
SPRIGMAN FINAL 12/17/2004 3:36 PM
interfere with the “enjoyment and exercise” of copyright. There are two senses
in which that phrase may be interpreted. The first relates to enjoyment and
exercise of the author’s economic rights. Under Article 9(1) of the Berne
Convention, authors of literary and artistic works have the exclusive right of
authorizing the reproduction of those works “in any manner or form.” This
includes traditional photocopying, digital copying, or any other form of
copying of the entire work or any part thereof. These exclusive rights include,
as a necessary corollary, the right to refuse to authorize reproduction of a
protected work. This right to exclude is the mainspring of the author’s
economic right: by restricting reproduction, the author may reduce output of his
work and thereby realize supracompetitive returns if his work lacks ready
substitutes.
Nevertheless, authors who fail to comply with new-style formalities and
thereby lose their previously existing right to exclude are likely not, as a
category, deprived of any aspect of the “enjoyment and exercise” of the
economic rights appertaining to their copyright. An author who fails to comply
with new-style formalities is merely converting an entitlement that is initially
protected by a property right (the right to exclude, realized through injunctions
and infringement damages) into an entitlement protected by a liability right (the
right to recover revenues from use via a default license).244 Even though new-
style formalities set up the liability rule as the default, and require authors to
opt out to preserve their ability to exploit the property rule, the system—unlike
the usual system of compulsory licenses—is still voluntary. And if the royalty
payable under the license is set correctly, owners of copyrights with projected
values lower than the cost of complying with a formality should actually prefer
the liability right as a means of exploiting their copyrights.
Creating an exploitation option based on a liability rule is a modest
extension of what the current regime provides. In the copyright system we have
now, rightsholders are allowed to choose the best approaches within the
existing set of property rules for exploiting their copyrights, whether via
exclusion and collection of infringement damages, a program of licensing and
collection of license royalties, or a mixture of the two. New-style formalities
would expand the existing process by establishing a liability rule option in the
form of a default license for works not valuable enough to justify customized
licensing. Again, whether the default license applies is within the control of the
rightsholder, and therefore a rightsholder’s decision to rely on that liability rule,
rather than on a property rule, to protect the enjoyment and exercise of his
copyright is not a forfeiture of rights. It is, rather, a signal that a particular
rightsholder believes that a one-size-fits-all liability rule based on a default
license is preferable, because of low transaction costs, to a property right
244. See Guido Calabresi & A. Douglas Melamed, Property Rules, Liability Rules, and
Inalienability: One View of the Cathedral, 85 HARV. L. REV. 1089, 1092-93 (1972).
SPRIGMAN FINAL 12/17/2004 3:36 PM
247. TRIPs Article 13 provides that “[m]embers shall confine limitations or exceptions
to exclusive rights to certain special cases which do not conflict with a normal exploitation
of the work and do not unreasonably prejudice the legitimate interests of the right holder.”
TRIPs, supra note 16, art. 13; see also Jane C. Ginsburg, Toward Supranational Copyright
Law? The WTO Panel Decision and the “Three-Step Test” for Copyright Exceptions, 187
REVUE INTERNATIONALE DU DROIT D’AUTEUR 3 (2001) (analyzing the meaning of Article
13).
248. RICKETSON, supra note 208, at 489.
SPRIGMAN FINAL 12/17/2004 3:36 PM
254. Id. ¶ 6.183; accord RICKETSON, supra note 208, at 483 (stating that “normal
exploitation” refers to “the ways in which an author might reasonably be expected to exploit
his work in the normal course of events”).
255. See generally 2 GOLDSTEIN, supra note 140, ch. 10 (explaining the fair use
doctrine).
256. Berne Convention, supra note 14, art. 10(1).
257. Id. art. 10(2).
258. Id. art. 10bis(2).
259. Even if default licenses did systematically interfere with rightsholders’ ability to
profit from their works, that interference remains permissible so long as it does not create
“unreasonable prejudice.” As the WTO panel made clear, whether an exception creates
“unreasonable prejudice” will depend in part on whether compensation is provided to
rightsholders: “in cases where there would be [a] serious loss of profit for the copyright
owner, the law should provide him with some compensation (a system of compulsory
licensing with equitable remuneration).” Panel Report, supra note 250, ¶ 6.229 n.205
SPRIGMAN FINAL 12/17/2004 3:36 PM
3% don't
100%
require
attribution
90%
50%
40%
97% require
30% attribution 67% don't allow
33% restrict commercial use
20% derivative works
10%
0%
The data set out in Figure 6—data which was provided by Creative
Commons—shows the choices that Creative Commons licensors have made
over the first two years of the organization’s existence regarding which rights
to reserve and which to give away. Although rightsholders who seek out, or
become informed about, Creative Commons and decide to enter into a Creative
Commons license are certainly not representative of rightsholders as a broader
group, the Creative Commons license distribution data gives us some insight
into what the world might look like when copyright is no longer an on/off
switch, but is more finely variegated.
Perhaps most unexpectedly, the data shows that a significant majority
(67%) of Creative Commons licensors allow the use of their content in the
creation of derivative works. This data suggests that many rightsholders would
voluntarily abandon control over derivative works, which is an element both of
the author’s economic right and his right of integrity—although most who do
so (again 67%) would limit use to noncommercial derivative works.
In contrast, the Creative Commons data shows that almost all licensors
(97%) require attribution in exchange for permission to use their works. That
number suggests that the norm favoring attribution is strong, and, consequently,
that we may succeed in moving copyright closer to rightsholders’ expectations
and simultaneously free a large amount of creative work if we install a
SPRIGMAN FINAL 12/17/2004 3:36 PM
For reasons that have been discussed earlier in this Article, we would
expect a large number of authors to fail to comply with registration and notice
requirements, and, similarly, the majority of rightsholders to fail to comply
with a renewal formality. Accordingly, a large number of works (in both
absolute and percentage terms) will be exposed to default licenses in a system
of new-style formalities. So one might read the WTO panel’s holding that
exceptions must be “narrow in a quantitative as well as a qualitative sense” as
ruling default licenses out as an exception permissible under Berne 9(2) or
TRIPs 13.
It should be noted, however, that the WTO panel’s report was issued in a
case involving royalty-free compulsory licensing, which means that, unlike in
the case of default licenses, the WTO panel was dealing with an exception that
eliminates rightsholders’ ability to receive compensation for their work for the
excepted uses. In that context, and in the absence of any particular provision in
Berne or TRIPs approving the particular compulsory licenses at issue, an
independent “special cases” element makes sense. If the touchstone of the
exceptions provision is preserving rightsholders’ abilities to profit from their
works, application of the “normal exploitation” and “legitimate interests”
elements might theoretically provide all the information needed to determine
whether an exception should be permitted. But, in practice, whether a particular
exception interferes with “normal exploitation” or a rightsholder’s “legitimate
interests” may be difficult to determine with certainty. The “certain special
cases” element can therefore be seen as providing an easily administered
threshold test: does this exception affect a large enough share of the potential
licensing market that it is likely to interfere with a rightsholder’s ability to
exploit his work?
But what role should the “certain special cases” element play in a case like
default licenses, where there is no systematic interference with rightsholders’
economic interests? In such an instance, the “certain special cases” language
may be both theoretically and practically coterminous with the “normal
exploitation” element, i.e., the number of works to which an exception applies
is required to be sufficiently limited that the market for a particular work is not
substantially affected. For works that have a ready market, even a relatively
small amount of excepted usage might affect the rightsholder’s ability to
exploit the work. But because the default licenses connected to new-style
formalities apply only after a rightsholder signals that his work does not enjoy a
substantial commercial market, even a relatively large number of excepted uses
made under a default license would not interfere with the rightsholder’s own
understanding of his ability to exploit the work. Put differently, there is a strong
argument that because default licenses arise only when the author or
rightsholder sends a signal that his work lacks significant commercial value, the
default license applies only in the “special case” where a liability rule is
preferred by the rightsholder.
SPRIGMAN FINAL 12/17/2004 3:36 PM
Perhaps the most that can be said at this point about the “special cases”
element—and indeed about the Berne Article 9(2)/TRIPs Article 13 test
altogether—is (1) that on a proper reading of Berne Article 5(2), the need to
justify new-style formalities as an “exception” will never arise, and (2) if the
exceptions provision is applicable, the elements of the test are sufficiently
indeterminate (at least as they apply to default licenses, a mechanism that the
Berne drafters could not have had in mind when they formulated the Article
9(2) test) that the application of the test will involve not so much testing new-
style formalities under the formal elements as an evaluation of whether the
system of new-style formalities serves (or undermines) the foundational
principles of the Berne Convention. If Berne is focused on protecting authors’
ability to exploit their works, new-style formalities will be compatible with the
Convention. If Berne is focused, instead, on enforcing a particular property
rule, regardless of authors’ interests, then new-style formalities may fail under
the current text of the Convention.
CONCLUSION